Our blog
What Does ChatGPT Know About Me? How to Check (and Fix) Your AI Answer
Type your own name into ChatGPT and you might get a flattering mini-biography, a confident description of someone who shares your name but isn't you, or a shrug: "I don't have specific information about that person." All three answers can happen in the same week, from the same assistant, and the reason isn't random. It comes down to how ChatGPT actually knows things — and that mechanism is worth understanding before you worry about what it's saying. This matters more than it used to. A growing number of people now ask an AI assistant "who is this person" before a job interview, a first date, a business deal, or a new hire — the same way people used to just search a name on Google. If the assistant's answer is outdated, mixed up with a stranger, or just incomplete, that's the impression someone forms of you, often without ever telling you they checked. Understanding why AI gives inconsistent answers, and what you can actually do about it, starts with the single most misunderstood part of how these tools work. Two different ways AI "knows" about you When people say "ChatGPT knows about me," they're usually describing one of two completely different processes, and mixing them up is the source of almost all the confusion. The first is training memory. ChatGPT, like Gemini and Claude, was built by having a model read an enormous amount of text — articles, forum posts, old web pages, directories — up to some cutoff point. If your name appeared in that material, the model may have absorbed a fuzzy impression of it. This memory is frozen. It doesn't update when you change jobs, move cities, or get a new article written about you. It can also be wrong in the first place, since training data includes outdated pages, satire, and other people's mistakes. The second is live web search. Modern versions of ChatGPT, Gemini, and Claude can, in some circumstances, actually search the current web before answering, the same way a search engine would, and summarize what it finds. This is far more current and often far more accurate — but it isn't triggered every time. Depending on the exact prompt, the product tier, and the assistant's own judgment about whether a search is "needed," you might get a stale, memory-based answer when a quick search would have given you something completely different. This single distinction explains most of what confuses people about AI and their own name: the same question, asked two different ways, can pull from two different sources and come back with two different pictures of you. Think of it like asking two different relatives about you. One hasn't seen you since a family reunion years ago and is working purely from memory — maybe accurate, maybe outdated, maybe mixing you up with a cousin. The other just called your current employer to check before answering. Both are trying to help; only one is looking at anything current. ChatGPT's training memory is the first relative. Its live web search, when it actually happens, is the second. The practical takeaway is that you can't judge "what AI thinks of me" from a single answer. A vague or wrong response might just mean the assistant stayed in memory mode and never looked anything up — not that there's necessarily something wrong being said about you across the web. If you want to see this in action rather than take it on faith, you can check what AI says about your name — free. Tailwind's tool asks several assistants directly and shows you the actual answers side by side, so you're not guessing which mode each one happened to be in. Why ChatGPT, Gemini, and Claude don't agree with each other Ask three assistants the same question about yourself and you may get three different answers — and none of them is necessarily "wrong" in the sense of being broken. They're just built differently. Different training data. Each company trained its model on a different slice of the internet, collected at a different time. One model may have picked up a local news mention of you that another never saw. Different recency. Training cutoffs land on different dates, so one assistant's "frozen" memory of you might be a year or two newer than another's. Different search behavior. Some assistants default to searching the live web more aggressively for "who is" style questions; others lean on internal memory unless you push them to look something up. Different weighting of sources. Even when two assistants search the same web, they can summarize it differently, giving more or less weight to a given article, profile, or forum thread. None of this means the answers are meaningless — it means a single check, on a single assistant, is only a partial picture. Someone with a very common name, like "David Chen" or "Maria Lopez," might see wildly different answers across the three because each model is latching onto a different David Chen or Maria Lopez entirely, not because any single model is malfunctioning. Someone with a more unusual name combination might see more consistent answers, simply because there's less for the models to confuse. This is also why it's worth checking more than one assistant if you can. Relying on whichever one you happen to have open gives you one version of a story that, by design, has at least three different tellers. How to actually check what AI says about you There are two practical ways to do this, and they're worth doing together rather than picking just one. The DIY option: just ask Open ChatGPT, Gemini, and Claude separately and ask each one a plain question like "What do you know about [your full name]?" or "Who is [your full name], [your city or profession] if that helps narrow it down?" Try it both with and without the extra identifying detail — the extra context sometimes nudges the assistant toward a live web search that a bare name alone would not trigger. If you have access to a version with browsing or search explicitly enabled, try asking it to "search the web for" your name directly, which tends to force the more current mode rather than leaving it to the model's judgment. Note what each one says, and whether it reads like a confident, current summary or a vague, hedged guess. A vague answer that leans on phrases like "I don't have specific information" usually means you're seeing training memory come up empty, not a search result. A confident but oddly specific answer about a job you've never had or a city you've never lived in usually means you're looking at someone else's information attached to your name. Repeat this across all three assistants and you'll start to see a pattern: which one has the most current information, which one tends to confuse you with someone else, and which one simply has nothing. That pattern is more informative than any single answer. The faster option: a dedicated name check Doing the above manually across three assistants, more than once, gets tedious fast — which is the whole reason Tailwind built a free AI name check. It runs your name through multiple AI assistants, shows you what each currently says side by side, and gives you a public profile page you can revisit later to see what's changed. There's no cost to check, and no account commitment required just to see the results. If you want to keep an eye on this over time rather than checking manually every few months, Tailwind also offers ongoing monitoring of your AI profile — either as a standalone add-on for a few dollars a month, or bundled into a fuller reputation plan. For a first look, though, the free check alone is enough to tell you where you stand. A concrete version of this: imagine Priya Nair, a project manager in Austin, asks ChatGPT who she is before sending her resume to a recruiter who she suspects will do the same search. The assistant describes a university lecturer with a similar name in a different country — plausible-sounding, entirely wrong, and potentially the first impression that recruiter forms before ever speaking with her. She didn't do anything wrong to end up here; she just happens to share a name with someone else who has more written about them. Knowing this is possible, and checking for it before it matters, is the whole point of running this kind of check proactively rather than finding out secondhand. What to do if the answer is wrong, confusing, or too revealing Once you actually see what's being said, you'll generally land in one of three situations — and each has a different next step. If the AI seems to be describing a different person entirely — wrong job, wrong city, wrong achievements — you're most likely dealing with a namesake collision, which is extremely common for anyone with a reasonably shared name. We cover why this happens and how to fix it in a separate guide on AI mixing you up with someone else. If the issue is bigger than AI — your phone number, home address, or other personal details are easy to find across the web in general, not just in AI answers — that's a broader internet exposure problem, and AI is really just surfacing it. Our guide to removing yourself from the internet walks through what's realistically possible to clean up. If the problem is specifically that an AI assistant's training data includes something you'd like reconsidered, there are official channels for submitting a request to the companies that build these models, though nothing about the process is instant or guaranteed. We go through what that process generally involves in our guide to requesting removal of personal data from ChatGPT. In practice, most people find they're dealing with a mix of these rather than just one. A namesake mix-up often sits on top of a thin, inconsistent online presence — the same gap that makes it easier for an AI model to blend you with someone else is often the gap that leaves your own accurate information absent from the answer entirely. Fixing one tends to help the other. Why does ChatGPT say it doesn't know anything about me? Usually one of two things: either you genuinely don't have much indexed, searchable presence tied clearly to your name, or the assistant answered from training memory and didn't search the live web at all. Try rephrasing with a city or profession attached, or explicitly ask it to search before answering. Can I make ChatGPT forget something about me right now? Not instantly, and not with certainty. There isn't a button anyone can press to surgically remove one fact from a trained model the way you'd delete a row from a spreadsheet. You can submit a request through the model provider's official process, and that can sometimes influence what shows up over time, but it isn't a guarantee and it isn't immediate. How often should I check what AI says about me? Every few months is reasonable for most people, more often if you're job hunting, dating, running a business under your own name, or you've recently had something written about you online. The answers genuinely do shift as the web changes and as these assistants update how often they search it, so a one-time check tells you about today, not about six months from now. The honest summary is that no single check gives you a permanent verdict. What you get is a snapshot — useful, often revealing, sometimes unsettling — of how a handful of AI systems currently represent you. Checking again periodically, especially after you've cleaned up anything misleading, is the only way to know whether the picture is actually improving.
How to Remove Yourself From the Internet: What Actually Works in 2026
Somewhere on the internet right now, there's a data broker site listing your old address, a forum post you forgot you wrote in 2014, and a news article that mentions you once in a context you'd rather not lead with. If you've ever searched your own name and felt a jolt of "wait, that's still up there," you've probably also wondered whether there's a way to just make it all go away. There isn't — and getting clear on that upfront is actually the most useful thing you can do before spending a single hour on this. That doesn't mean the effort is pointless. It means the project is different from what the phrase "remove yourself from the internet" implies. Think of it less like deleting a file and more like cleaning up a cluttered house you still have to live in: you can get rid of a genuine majority of the mess, box up and dispose of the parts that are easiest to control, and make sure the rooms people actually walk through look the way you want them to. What you can't do is pretend the house was never lived in. This guide walks through what realistically moves the needle, what doesn't, and why checking what AI assistants say about you has become part of the same project as checking a search engine. Why full erasure isn't realistic — and what is You cannot fully remove yourself from the internet. Not with a paid service, not with a weekend of diligent form-filling, not with a lawyer. A few things make total erasure structurally impossible, not just difficult: Cached and archived copies. Search engines, the Wayback Machine, and countless scrapers keep snapshots of pages independent of whether the original page still exists or changes. Other people's posts about you. A photo a friend tagged you in, a mention in a local news story, a comment on someone else's thread — you don't control accounts you don't own, and most platforms won't remove content just because it names you. Public records. Property records, court filings, business registrations, and voter files are, in most places, public by design, and they regularly feed the people-search sites that frustrate people the most. Re-scraping data brokers. Even after a successful opt-out, many broker sites rebuild their databases from the same public sources on a regular cycle, which means a listing you removed can simply reappear under a new URL months later. So the realistic goal isn't "erase everything" — it's "remove what you can control, and outweigh what you can't." That second half matters as much as the first. Once you've stripped out the easy, embarrassing, or dangerous exposure, the next job is making sure the information that remains findable about you is accurate and current, rather than leaving the field to a ten-year-old forum post or a broker listing with your old address. Removal and replacement are really one project, not two. Consider David Chen, a consultant who spent a weekend opting out of a dozen data broker sites after a client mentioned finding his old home address online. The opt-outs worked — the broker listings came down within a few weeks. But a five-year-old local news item about an unrelated legal dispute involving someone else named David Chen still ranked on the first page for his name, simply because nothing more current and specific to him outranked it. Removing the broker listings was necessary. It wasn't sufficient on its own, because it didn't do anything to push an accurate, current picture of him higher. That second part — outweighing what you can't remove — is where a lot of people stop too early. What actually works, roughly in priority order Not every tactic here is worth equal time. If you only have an afternoon, start at the top of this list — it's ordered by how much exposure it typically removes per hour spent. 1. Data broker opt-outs People-search sites — the ones that show your address, age, phone number, and "possible relatives" when someone searches your name — are usually the biggest single source of exposed personal information, and they're also the most directly fixable. Most of these sites have an opt-out process, though it's typically a manual, repetitive one spread across dozens of individual sites rather than a single switch: you find your listing, submit a removal request (sometimes by form, sometimes by email), and often wait anywhere from a few days to a few weeks for it to actually disappear. The frustrating part is that this isn't a one-time task. Because these sites rebuild their databases from public records on a recurring basis, a listing you successfully removed can reappear months later under a slightly different URL, with the same information. This is less a sign you did something wrong and more a structural feature of how the data broker industry operates — which is exactly why some people decide it's worth paying a service to handle the repetition for them, a tradeoff we cover further down. This is big enough as its own topic that we've written a dedicated, detailed list of sites and how to opt out of each — see our full data broker opt-out guide if this is where you want to start. 2. Google's tool for removing personal info from search results Separate from the broker sites themselves, Google offers a way to request removal of certain personal information — like a phone number or home address — from its own search results, even if the underlying page stays up elsewhere. It's a narrower tool than people expect: it affects what shows up in Google search, not the source page, and not other search engines. Still, since most people's "checking up on someone" starts with a Google search, this is a high-leverage, low-effort step. We walk through exactly how to use it in this guide to Google's results-about-you tool. 3. Old social media accounts and abandoned profiles That Myspace account, the Tumblr you forgot existed, a half-finished dating profile from years ago — these sit around collecting search visibility and sometimes outdated personal details long after you've stopped thinking about them. Go through old email accounts for "welcome to" and "verify your account" messages to find profiles you've forgotten about, then either delete them outright or lock them down as far as the platform allows. Deletion is almost always better than deactivation, since deactivated accounts can sometimes still be found or reactivated, while deleted ones are generally gone for good. It's worth doing this even for accounts that feel harmless. An old, abandoned profile with a stale profile picture and an outdated job title doesn't just sit there quietly — it's exactly the kind of thin, inconsistent content that a search engine or an AI assistant can end up summarizing as part of "who you are," simply because it's one of the few concrete data points attached to your name. Cleaning it up isn't just tidiness; it removes a source of inaccurate information before it has a chance to get picked up anywhere else. 4. Direct removal requests to specific sites For a site that republishes your personal information outside the major broker networks — a resume aggregator, an old forum profile, a defunct company's "team" page that still lists you — a direct, polite removal request to the site owner or its webmaster often works, especially for small or low-traffic sites with no real reason to keep your information up. Keep the request specific: identify the exact page, what it shows, and why you're asking it to come down. 5. Outdated news and directory listings This is the hardest category, and it's worth being honest about that rather than promising more than is realistic. Legitimate news organizations are generally under no obligation to remove accurate reporting just because it's inconvenient for you later, and most won't take down a factual article on request. That said, some outlets will update, correct, or add context to an outdated piece if you reach out respectfully and explain what's inaccurate or outdated — it's worth asking, just without the expectation that it will work. Directory sites and aggregators that scraped the same story are a separate, easier target, and often respond to the same kind of direct removal request as in the previous step. Old directory listings tend to be the quiet, low-stakes version of this problem: a law firm's "former associates" page that still lists you years after you left, a university alumni directory with your old email, a conference speaker-bio page from a job you no longer hold. None of these are damaging exactly, but they add up to a search-results page that looks out of date, which has a way of making everything else about you look slightly less current too. A short email to the site owner, asking them to update or remove the listing, resolves a surprising share of these without any pushback. What paid removal services actually do — and don't Services like DeleteMe, Incogni, and Aura exist because the data broker opt-out process described above is genuinely tedious: there are dozens of broker sites, each with its own opt-out form, and most require you to repeat the process periodically as listings reappear. These services automate exactly that — submitting and re-submitting opt-out requests across a wide list of data broker sites on an ongoing basis, so you don't have to do it by hand every few months. That's a real, useful service, and it's reasonable to pay for convenience rather than spend your own evenings on it. Pricing for these services typically runs as an annual subscription, though exact plans and costs vary by provider and change over time. The three names people most commonly compare are DeleteMe, Incogni, and Aura. In broad terms, DeleteMe has historically positioned itself around broker opt-outs with some manual review by its team, Incogni leans toward a more automated, lower-touch approach to the same core task, and Aura tends to bundle broker removal together with other protections like identity-theft monitoring rather than selling it as a standalone product. The differences matter less than the shared ceiling: all three are, at their core, automating repeated opt-out submissions to data broker sites, not reaching into the parts of the internet that sit outside that category. What they can't do is just as important to understand before you pay for one. They can't remove a photo your cousin posted of you, or a comment you left on a news article under your real name, or any other content someone else controls. They can't force a search engine to instantly clear its cache the moment a broker listing comes down. And they have no mechanism at all for AI training data — no paid service can reach into a language model and edit what it learned during training, because that isn't how these systems work; removal services operate on data broker sites and some search-result processes, not on model weights. If you're comparing these services to decide whether one is worth paying for, our deeper comparison of data removal services goes through the specifics of what each one covers. Once you've done the manual and automated cleanup, it's worth checking the thing none of these services touch at all: what AI assistants currently say if someone just asks them about you by name, which is a separate surface from search results or broker listings. You can see what AI still says after the cleanup for free — it's a useful final check specifically because it's the one place none of the removal work above actually reaches. How much of this should I actually expect to work? Expect partial, meaningful progress rather than a clean slate. Data broker opt-outs reliably reduce how much shows up when someone searches your name casually, even though specific listings can reappear over time. Google's personal-info removal tool reliably narrows what shows up in search for the specific items it covers. Old accounts, once deleted, generally stay deleted. Direct requests to small sites succeed more often than people expect, mostly because there was no strong reason for the information to be there in the first place. News coverage and major public records are the parts that typically stay, and that's a realistic outcome to plan around rather than a failure of the process. What should I do first if I only have an hour? Search your own name in quotes and screenshot or list every result that bothers you, then sort them into the categories above: broker listing, old account, specific site, or news/directory piece. That sorting is most of the value of the first hour, because it tells you which of the tactics above actually applies to your situation instead of guessing. From there, broker opt-outs and the Google tool are usually the best return on a single hour, since both tend to affect multiple results at once rather than one at a time. Will any of this guarantee nothing bad ever comes up about me again? No, and it's worth being direct about that rather than selling false certainty. New data gets generated constantly — a new public record, a new mention, a re-scraped broker listing — so there's no finish line where your exposure is permanently zero. The honest goal is reducing your exposure meaningfully and keeping an eye on what remains, not reaching a state of total invisibility. Should I do this myself or pay for a service? It depends mostly on your time and how many broker sites you're dealing with. If you're comfortable doing a focused weekend project and don't mind repeating parts of it every few months, doing it yourself costs nothing but time. If you'd rather not track dozens of individual opt-out forms and their renewal cycles, a paid service trades money for that ongoing effort — just go in understanding what it will and won't reach, as covered above. How long does this whole process take? The first pass — opt-outs, the Google tool, deleting old accounts — is realistically a few focused sessions rather than a single afternoon, since there are a lot of individual sites and forms involved. After that, it shifts from a one-time project into light maintenance: periodic re-checks every few months, since broker sites rebuild their data and new mentions of you appear over time regardless of how thorough your first pass was. That maintenance mindset is really the point of this whole guide. Nobody finishes "removing themselves from the internet" the way you'd finish cleaning a garage. You do the real work of cutting down what's easily findable, you make sure what's left and what AI assistants repeat about you is accurate rather than outdated or wrong, and then you check back periodically, because the web — and what's trained on it — never actually stops changing. A reasonable rhythm for most people is a check every few months: a quick search of your own name, a glance at whether any new broker listings have popped back up, and now, increasingly, a quick look at what an AI assistant says if someone simply asks it who you are. None of these take long individually. Done occasionally and consistently, they do more for how you show up online than any single big push ever will.
Google Results About You: How to Remove Your Phone Number and Address From Search
Type your own name into Google, in quotes, and look past the first couple of results. For most people, somewhere on that first page or the one after it, there's a listing from a site they've never visited and never signed up for — showing a phone number, a current or past home address, an age, and a list of "possible relatives" that reads like a family tree assembled by a stranger. It's unsettling the first time you see it, mostly because you never agreed to any of it. These listings usually come from people-search sites and data brokers that scrape public records — voter files, property deeds, court filings — and repackage them into a searchable profile, then optimize that profile to rank well for your exact name. Take someone like Maria Lopez, a dental hygienist who searched her own name after a coworker mentioned finding her cell number online in under a minute. The first result was her clinic's staff page — fine, expected. The second and third were people-search listings showing her current address, her age, a previous address from six years earlier, and the names of two relatives. She hadn't posted any of that herself. It had been pulled together from public records and published by sites whose entire business model is making that kind of lookup fast and free for anyone who searches. The good news is that Google gives you a direct way to ask for this specific kind of result to be removed from search — separate from contacting the site itself. The less good news is that it only solves part of the problem, and it's worth being clear-eyed about which part before you spend an evening on it. Once you've handled your search results, it's also worth taking the newer step of checking what AI assistants currently say if someone asks them about you by name — check what AI repeats about you for free, since that's a separate question from what shows up in a Google search. What Google's "results about you" process actually covers Google has a dedicated process for requesting the removal of search results that expose personal contact information — things like your phone number, home address, or email address, especially when that information appears alongside your name on a site you don't control, such as a people-search or data-broker listing. It's built around a simple idea: Google can't police every data broker on the internet, but it can control what its own search engine chooses to surface, and it has decided that this category of personal exposure is worth a dedicated removal path rather than making people file a generic complaint. To use it, you generally submit the specific web page URLs that contain your information, along with the search terms that surface them, through a form built for this purpose. Google then reviews each submission against its policy for personal information removal and decides whether the page qualifies — not every page that mentions you will, and a page that's clearly newsworthy, part of the public record, or already widely published elsewhere is less likely to qualify even if it feels invasive to you. What counts as eligible, in general terms Google's stated policy is aimed at things like phone numbers, home addresses, personal email addresses, and similarly sensitive identifiers, along with a few other categories such as certain ID numbers or images used in doxxing contexts. It is not a general-purpose tool for removing unflattering news coverage, old social media posts, or content you simply don't like. If a result embarrasses you but doesn't expose contact information, this particular tool usually isn't the right lever — that's a different, harder conversation about reputation rather than privacy. The part most people miss: this doesn't touch the page itself Here's the detail that trips people up, and it matters enough to say plainly: even a fully successful request through this process removes the result from Google's search listings. It does not delete the underlying page. The data-broker site that published your phone number and address is still live, still has your information, and can still be found by someone who navigates to it directly, finds it through a different search engine, or stumbles onto it through a link on another site. You've made it harder to find by searching your name on Google — you haven't made the information disappear. That distinction is worth sitting with for a second, because it changes what "done" looks like. A thorough approach usually means two separate actions: asking Google to stop surfacing the result, and separately contacting the site that published it to request they remove or opt you out of the listing (most data brokers have their own opt-out process, often buried a few clicks deep in their privacy policy). Doing only the first one leaves the actual source untouched. There's a second blind spot worth naming explicitly: none of this affects what an AI assistant says if someone asks it directly about you in a conversation. Removing a search result changes what shows up on a results page — it doesn't reach into a model's training data or retrieval layer and edit what it says when prompted with your name. Those are genuinely separate systems, checked and addressed separately. Search removal and AI-answer removal are not the same project, even though they both start with the same anxious impulse to type your own name somewhere and see what comes back. What to gather before you submit a request The single biggest mistake people make with this process is submitting a vague request — "please remove results about me" — instead of a specific one. Google's review process works off exact URLs, not general complaints, so before you start, spend fifteen minutes doing the following: Search your exact name in quotes and go through several pages of results, not just the first one, noting every URL that shows your phone number, address, or other contact details. Try variations — with and without a middle name, with a former last name if applicable, and with common misspellings, since data brokers often index several versions of the same person. Open each result and confirm the information shown is actually yours and actually current — people-search sites frequently display outdated addresses or mismatch records between people with similar names. Copy the exact URL of each page, not just the domain — a generic request pointing at a site's homepage is far weaker than one pointing at the specific profile page showing your details. Note which search terms surface each result, since the submission form typically asks for both the URL and the query that returned it. Having this list ready before you open the form makes the whole process faster and gives Google's reviewers something concrete to act on, rather than a general request they have to interpret. It also helps to keep a simple record — a spreadsheet row per URL, with the date you submitted and the outcome — since you may end up repeating parts of this process down the line, and it's much easier to pick up where you left off than to start from scratch trying to remember which listings you already dealt with. Realistic expectations on timing and outcomes This is not an instant process, and it's not a guaranteed one. Submissions get reviewed individually, and review times can vary — plan on this taking some real time rather than expecting same-day results. Some requests are approved as submitted, some are partially approved, and some are declined because the page doesn't meet the policy's bar — for instance, if the information is also published somewhere clearly newsworthy, or if the page doesn't actually contain the category of personal information the policy covers. It's also worth knowing that approval isn't permanent protection. Data brokers frequently republish removed listings, sometimes under a slightly different URL, which can mean the same information resurfaces later and requires a fresh request. This isn't a sign the process failed — it's just how an unregulated industry of scraped-data sites tends to behave, and it's part of why people who are serious about this treat it as an ongoing check-in rather than a one-time errand. Where this fits in the bigger picture Submitting one removal request can feel like progress, and it is — but it's one piece of a larger project if your actual goal is reducing how much of your personal information is easy to find online. That larger project usually involves identifying which data brokers have listed you, requesting opt-outs from each one directly, cleaning up old accounts and public profiles, and deciding how to handle anything that's genuinely difficult to touch, like court records or old news coverage. For the fuller picture of what that looks like end to end, this guide to removing yourself from the internet walks through it step by step. And once you've worked through your search results, it's worth doing the newer version of the same check: asking what AI assistants currently say about you if someone just asks them your name in conversation, rather than searching for it. That's a different surface with a different fix, and increasingly, it's the one people encounter first — before they ever open a search engine at all. Does removing a search result also remove it from AI assistants? No. Google's removal process changes what appears in Google's own search results. It has no effect on what ChatGPT, Gemini, Claude, or any other AI assistant says when asked about you directly, since those systems draw on their own training data and, in some cases, their own live retrieval — not on Google's search index in real time. If you want to know what an AI assistant currently says about you, you have to check that separately. If my request gets approved, am I done? Approved means the specific result you flagged has been removed from that search query on Google. It doesn't mean the source page is gone, and it doesn't mean a similar listing won't appear later, sometimes from the same broker under a new URL. Treat it as one completed task in an ongoing process rather than a final fix.
How to Google Yourself the Right Way (and What to Look For)
Most people's version of "Googling myself" is a quick, half-distracted glance — type the name, skim the first few results, close the tab. It takes thirty seconds and tells you almost nothing useful, partly because the results you see are tailored to you specifically, based on your search history, your location, and your past clicks. A stranger searching your name sees something different. If you actually want to know what's findable about you, it's worth doing this properly, once, and then treating it as something you repeat occasionally rather than a one-time curiosity check. Step one: search like a stranger would Open an incognito or private browsing window before you start. This matters more than it sounds like it should — a regular, logged-in search reflects your personalization, which can bury results a stranger would see clearly, or surface things only because you've clicked on similar content before. Private mode gets you closer to a neutral, generic version of your results. From there, work through a few passes rather than a single search: Search your exact name in quotation marks — "Maria Lopez" rather than Maria Lopez — which tells Google to look for that exact phrase rather than loosely matching the words anywhere on a page. Check the Images tab. This surfaces photos tagged or associated with your name that might not appear in a standard web search, including ones from sites you don't remember being on. Check the News tab, especially if you have a reasonably distinctive name or have ever been quoted, mentioned, or involved in anything that made local or industry press. Try variations — a maiden name, a nickname you used to go by, a middle name included or dropped, and a couple of common misspellings. People searching for you don't always know the "correct" version of your name, and neither does everyone who's published something about you. Add context terms — your city, your employer, your profession — paired with your name. This surfaces a different slice of results than your name alone, and it's often closer to how someone who actually knows a little about you would search. Going through all of this takes maybe fifteen minutes, not thirty seconds, but it gets you an actual picture instead of a glance. Once you've done the thorough version, there's a natural next step worth taking immediately after, while your own results are still fresh in your mind: then ask ChatGPT and Gemini too. A Google search only shows you Google's index. It says nothing about what an AI assistant would tell someone who simply asked it about you directly instead of searching — which is a different system, built on different data, and worth checking separately, either by asking each assistant yourself or through a free tool built to check all three at once. What to actually look for once you've done this Running the searches is the easy part. Knowing what you're looking at takes a little more attention. A few categories worth flagging as you go: Outdated information An old job title, a former employer, a city you haven't lived in for years, a phone number tied to a plan you canceled long ago — this kind of stale information is extremely common and usually harmless on its own, but it adds up. If someone's forming a first impression of you from search results, outdated context can paint a picture that's technically true at some point in the past and misleading now. Pay particular attention to profiles you set up once and never touched again — an old conference bio, a directory listing from a job you left years ago, a speaker page for an event that's long over. These tend to rank reasonably well precisely because they're old and have had time to accumulate links and search signal, which means an outdated version of you can genuinely outrank the current, accurate one if you haven't kept your active profiles updated. Things you'd rather weren't the first result This doesn't have to mean anything scandalous. It might be an old blog post you wrote in a very different voice than you'd use today, a social account you forgot still exists, a local news mention tied to something minor, or a review you left somewhere that's more visible than you'd expect. None of this is necessarily a crisis — it's just useful to know it's there before someone else finds it first. A namesake's content mixed in with yours If you have a reasonably common name, there's a good chance at least some of what comes up isn't actually you — it's someone else who shares your name, sometimes in a context wildly different from your own (a different profession, a different country, occasionally something you'd really rather not be confused with). This is worth identifying clearly, not because you can necessarily fix it, but because it helps you understand what a stranger doing a quick search might mistakenly attribute to you. Take David Chen, a structural engineer who ran this exact check and found his own LinkedIn and firm bio sitting comfortably in the top results — exactly what he'd hope for. Lower down the page, though, was a court records listing for a different David Chen entirely, in a different state, tied to a civil dispute that had nothing to do with him. Nothing about the listing claimed to be him specifically, but a hurried reader skimming results probably wouldn't notice the different middle initial and city. He couldn't get that page removed — it wasn't about him, so there was nothing to request — but knowing it was there meant he could add a few more identifying details to his own profiles (his specific city, his licensing credentials) to make the distinction clearer for anyone comparing the two. Personal information that shouldn't be as visible as it is While you're going through results, keep an eye out for your own phone number, home address, or other contact details showing up through a people-search or data-broker listing. This is a common enough find that it deserves its own separate response rather than getting lumped in with the "things I'd rather weren't there" category — there's a dedicated process for requesting this kind of personal information be removed from search results specifically, which is worth tackling as its own project once you've finished this broader pass. The limitation of this whole exercise Here's the part that's easy to miss, and it's becoming more relevant every year: a thorough Google search only shows you what Google's search index has decided to surface. It tells you nothing about what an AI assistant would say if someone simply asked it about you in conversation — "who is [your name]," "tell me about [your name] who works in [your field]" — instead of running a search and clicking through results themselves. That distinction matters more than it used to, because asking an AI assistant directly is an increasingly common way people actually get information now, especially for quick background checks — before a first date, before a meeting, before extending someone trust. Someone forming an impression of you this way never sees a results page at all. They see a generated answer, assembled from whatever the model learned during training and, in some cases, from information it retrieves live. That answer can be accurate, outdated, thin, confused with a namesake, or occasionally just wrong — and a Google search, no matter how carefully you run it, won't tell you which of those is happening. It's also a fundamentally different kind of exposure. A search result is something you can click into, read in context, and judge for yourself — you can see the source, the date, the surrounding page. An AI assistant's answer about you arrives as a single, confident paragraph with no visible sourcing and no easy way to check where any particular detail came from. If it's slightly wrong, or mixing you up with someone else who shares your name, there's no obvious way to notice unless you specifically go looking, the way you'd specifically go looking by running a search. So is Googling myself still worth doing? Yes — it's just not the complete picture anymore. Search results are still what shows up if someone looks you up by clicking around rather than asking an AI assistant directly, and plenty of people still search the old-fashioned way. A careful search also catches things an AI answer might not mention at all, like specific pages, photos, or namesake confusion that's genuinely useful to know about. It's a necessary first pass — just not the last one. What's the fastest way to check the AI side too? Asking ChatGPT, Gemini, and Claude each individually — "what do you know about [your name]" — works, but doing it three separate times in three separate apps is tedious enough that most people never get around to it. Running the same question across all three at once, in one pass, right after you've finished going through your actual search results, is the faster version of the same check. Between the two, you end up with a genuinely complete picture: what shows up when someone searches for you, and what gets said when someone simply asks about you instead. For a deeper look specifically at the AI side of that picture, the AI visibility checker is built for exactly that comparison — treat it as the second half of the exercise you just did, not a replacement for it.
Personal Reputation Management: A Do-It-Yourself Guide
Search "reputation management" and almost everything that comes up is written for businesses — how to get more five-star reviews, how to respond to an angry customer, how to protect a brand's name. That's a different problem from the one most individuals actually have: managing what shows up when someone searches your own name, not a company's. This guide is for that second case — a practical, do-it-yourself approach to your personal reputation, with an honest section on when DIY genuinely isn't enough. Personal reputation management is a different problem than business reputation management A business has official channels it controls — a website, a verified listing, a social account with a blue checkmark — and a reasonably clear strategy of soliciting reviews and responding to them publicly. A person usually doesn't have anything that organized. Your online presence is probably a scattered mix of old social profiles, a LinkedIn you update once a year, maybe a mention in a local news article from 2016, a few photos someone else tagged you in, and now, increasingly, whatever an AI assistant says when someone asks it about you by name. There's no single dashboard for any of that, and no "respond to this review" button. The goal isn't to chase down and fix every scattered mention — it's to make sure the handful of sources people are actually likely to find tell an accurate story. What you can realistically do yourself Build one or two strong, accurate anchor profiles Instead of trying to control every scattered mention of your name, put your energy into one or two profiles that are genuinely strong and that you keep accurate — typically a LinkedIn profile and, if it's relevant to your field, a simple personal website or portfolio. Search results tend to favor well-established, regularly updated profiles with real engagement, so a strong anchor profile often outranks older, messier, half-forgotten pages simply by being more complete and more current. Consider a fictional example: Priya Nair, a freelance designer, had an outdated Behance portfolio from an old job, a sparse LinkedIn, and a personal site she hadn't touched in three years. Consolidating her best, current work onto one updated LinkedIn profile and one simple personal site — and updating both regularly — gave search engines a clear, accurate, current signal to surface instead of a scattered mix of old, half-maintained ones. Keep your name format consistent everywhere If you go by "Maria Lopez" on LinkedIn, "Maria R. Lopez" on your personal site, and "Mari Lopez" on an old portfolio, you're fragmenting your own search signal across variations instead of consolidating it. Picking one consistent format and using it everywhere you have control makes it easier for the profiles you actually want found to rank together, and makes it easier for you to notice when something's missing or wrong. Fix factually wrong information at the source If you find something concretely wrong — an old job title, a city you haven't lived in for years, a professional affiliation that ended — go to where it originates and correct it there, rather than only trying to push it down in search results. An outdated LinkedIn connection's company directory, an alumni page, an old staff bio still live on a former employer's site — these are often editable or correctable with a direct, polite request to whoever maintains the page. Fixing the source is more durable than trying to outrank it, because the wrong information stops spreading further once it's corrected where it lives. Be visible in legitimate, verifiable ways If part of what's bothering you is that there isn't much accurate, positive information about you findable at all — rather than there being something specifically negative to counter — building a stronger presence through things that are real and verifiable (speaking at an industry event, writing articles in your field, doing an interview or podcast appearance, publishing work publicly) does more good than trying to suppress or bury anything. This only works because it's genuine; a stronger online presence is a byproduct of visible real activity, not something you can manufacture by posting keyword-stuffed content with no substance behind it. When DIY genuinely isn't enough There's no shame in reaching the edge of what a personal checklist can fix. A few situations are specifically beyond DIY, and recognizing them early matters: Ongoing harassment or a coordinated attack — a single rude comment is different from a sustained, organized effort across multiple platforms to damage your name. That pattern usually needs platform reporting tools used systematically, and sometimes outside help to manage the volume and the toll it takes. Doxxing — someone has published your address, phone number, or other private identifying information with intent to intimidate or expose you. This is a safety issue first, not just a reputation issue, and it often warrants both platform reporting and, depending on severity, law enforcement or legal involvement. Defamation — someone has published a false statement of fact about you that's causing real harm, not just an opinion you dislike. Whether something legally qualifies as defamation is a specific legal question with real nuance, and it's not something a blog post can responsibly answer for your specific case. In any of these, the right move is a lawyer who handles online harassment, privacy, or defamation, or a specialized service built for exactly this kind of situation — not pushing through with search-result tactics that were never designed to handle an active safety or legal issue. None of this is legal advice; it's a signal for when to go get some. Reputation management agencies aren't the enemy It's worth saying plainly: legitimate reputation management agencies exist, and they do real, skilled work for genuinely complex cases — coordinated harassment, serious defamation, situations with real legal and safety dimensions where DIY tactics clearly aren't enough. This guide isn't an argument against ever using one. It's written for the much more common situation: you don't have an active crisis, you just want the handful of things people are actually likely to find about you to be accurate and current. For that case, a paid agency is usually more than the problem calls for, and the steps above get you most of the way there yourself. How do I know if my situation needs professional help? A rough gut check: if the issue is "my information is outdated or scattered," that's DIY territory. If the issue involves ongoing targeted harassment, safety concerns, or something you suspect might be legally actionable, that's a sign to get a professional opinion — a lawyer, specifically, rather than a general reputation agency, if there's a legal question involved. Is it worth paying for a "remove my information" service for routine cleanup? For routine, non-urgent cleanup — old profiles, outdated listings, basic data broker exposure — many people can do the bulk of it themselves with some patience, though it does take real time. Whether a paid service is worth it for you usually comes down to how much your own time is worth and how large the cleanup project actually is. This breakdown of whether data removal services are worth it walks through that trade-off in more detail. The AI dimension is now part of this Everything above is the traditional version of personal reputation management — search engines, social profiles, old mentions. But your reputation today also includes a newer layer: what ChatGPT, Gemini, and Claude say when someone asks them about you directly, by name. That's a genuinely different surface than a search results page. It's conversational, it's framed as an answer rather than a list of links, and most people have never once checked what it currently says about them. The same principles apply here as everywhere else in this guide — you can't control it the way you'd want to, but you can know what it says, catch anything clearly wrong early, and address what's realistically addressable. Treating an occasional check of what AI says about you as part of routine personal reputation management, the same way you'd periodically Google yourself, is a small habit that costs almost nothing and tells you something you otherwise have no way to find out. Tailwind's free personal reputation tool is built for exactly that — a simple way to monitor your personal reputation in AI search without having to interrogate three different chatbots by hand every time you want to check. None of this requires turning into a full-time project manager of your own name. A couple of strong, accurate anchor profiles, consistent naming, fixing what's concretely wrong at the source, genuine visibility if you want more of it, a clear sense of when to call in a professional instead of pushing through alone, and an occasional check on what AI assistants are saying — that's a realistic, sustainable version of managing your own reputation, without pretending you can control everything people might ever find.
Are Data Removal Services Worth It? DeleteMe vs Incogni vs Doing It Yourself
David Chen typed his own name into Google one evening after a friend forwarded him a link to a people-search site with his phone number, home address, and a list of relatives attached. He spent twenty minutes clicking through opt-out forms on three different broker sites before he found the ads for DeleteMe and Incogni promising to do the whole thing for him. The question he actually had — and the one most people land on eventually — wasn't "which one is best," it was simpler: is paying someone else to do this actually worth it, or is it something I should just do myself? What these services actually do Strip away the marketing and the core function is straightforward: these are automation services for a manual chore that already existed. Data broker and people-search sites — the ones that scrape voter rolls, property records, and court filings into searchable profiles — almost all have an opt-out process. It's usually free, and it's usually tedious: find the listing, locate the opt-out form (often buried several pages deep), submit a request, sometimes verify your identity, and wait. Multiply that by the dozens of broker sites that exist, and the chore becomes a multi-hour project rather than a five-minute one. A removal service essentially runs that same process on your behalf, at scale, and repeats it periodically. You give them your name, address history, and sometimes other identifying details; they submit opt-out requests across a list of broker sites they track, and they re-check periodically because brokers have a habit of re-scraping public records and re-listing people who were previously removed. The pitch isn't "we can do something you can't" — it's "we'll do the thing you could do yourself, but keep doing it, so you don't have to remember to check back every few months." DeleteMe, Incogni, and Aura: how they generally differ All three operate on the same basic model — submit your information, they file opt-outs across a broker list, they send you periodic reports — but they differ in emphasis and approach. DeleteMe DeleteMe has been in this space longer than most competitors and leans on human review alongside automation — opt-out requests are reportedly handled by a mix of software and people checking submissions, rather than a fully automated pipeline. It's typically positioned as a subscription with family or multi-person plan options, and its reporting tends to include a readable summary of what was found and what was removed. Pricing has historically been on the higher end of this category. Incogni Incogni, from the company behind the Surfshark VPN brand, leans more heavily automated and markets itself around covering a broad list of broker sites with less manual back-and-forth. It's generally positioned as a lower-cost entry point into this category compared to some competitors. Its reporting dashboard tends to show removal progress as a running tally rather than a narrative report. Aura Aura bundles data broker removal into a wider identity-protection product — credit monitoring, identity theft insurance, and device security are typically sold alongside it rather than as a standalone opt-out service. That makes it a different kind of purchase decision: you're not just buying broker opt-outs, you're buying a broader security suite that happens to include them, which may or may not match what you're actually looking for. None of these differences should be taken as settled fact without checking current details — pricing tiers, the exact number of brokers covered, and specific features all change often enough in this category that anything you read, including this article, should be verified against the provider's current site before you decide. A few practical questions are worth asking of any provider before you sign up, regardless of brand: How many broker sites do they actually monitor, and is that list public or just a vague number in marketing copy? How often do they re-scan for new listings — monthly, quarterly? Do they send you a report you can actually read and verify, showing which specific sites were checked and what was found, or just a generic "we're working on it" dashboard? And can you cancel easily, or does the subscription require a phone call and a retention pitch? None of these questions have a universally right answer, but a provider that answers them clearly and specifically is generally more trustworthy than one that stays vague. Doing it yourself: what the manual version actually looks like If you'd rather skip the subscription, the manual process is not complicated — it's just repetitive. Here's roughly what it involves: Find out where you're listed. Search your name along with your city or a past address to surface people-search sites. A handful of names come up repeatedly across this category — the same sites tend to show up for most people, since they draw from the same underlying public records. Locate each site's opt-out page. This is usually linked from the footer or buried in the privacy policy, often under a phrase like "do not sell my information" or "opt out." Some are a simple form; others require you to find your own listing first and copy a profile URL into the request. Submit the request and verify your identity if asked. Some sites ask for an email confirmation, some ask for nothing beyond the listing URL, and a few have you upload identification — which, fairly, some people are uncomfortable doing on a site they don't trust. It's reasonable to skip any site that asks for more verification than you're comfortable providing. Track what you've submitted and when. A simple spreadsheet — site name, date submitted, date confirmed — saves you from re-doing work or losing track of which sites actually responded. Check back periodically. This is the step people skip, and it's the step a paid service is mostly selling you out of having to remember. Brokers re-scrape public records on their own schedule, so a site that removed you in January can show you again by summer. Realistically, the first pass through this takes a few hours spread across an evening or two, and the ongoing maintenance is lighter — mostly just re-checking every few months. For many people, that's a completely reasonable trade against a monthly subscription fee. For people who know they won't follow through on step five, paying someone else to remember it is arguably the better use of the money. If you decide to go this route, it helps to work from an actual list of broker sites rather than searching for each one individually — the slower part of the manual process is usually just finding where to start, not filling out the forms themselves. What none of them do — and this part matters more than the comparison above This is the section worth reading slowly, because the marketing around this category tends to blur past it. Regardless of which service you pick, or whether you do it yourself, there are things none of them can touch: They can't remove things other people post about you. A tagged photo, a mention in someone else's public post, a comment thread, a news article quoting you — none of that is a data broker listing, and opt-out automation has no mechanism to reach it. They can't instantly clear search engine caches. Even after a broker removes a listing, a cached version can linger in search results for a while, and no removal service controls how fast a search engine's index updates. They can't touch what an AI assistant learned during training. If your name, old address, or other details were part of the data a large language model was trained on, there is no opt-out form that reaches into that training data and edits it. This is a fundamentally different kind of data from a broker's live, re-scrapable database, and no company — not a removal service, not even the AI companies themselves acting on a request — can promise surgical, on-demand deletion from a trained model. They can't guarantee a broker stays removed. Opt-outs reduce how often you're re-listed; they don't permanently disable the business model of sites built on re-scraping public records. That's exactly why these services are subscriptions rather than one-time purchases — the work is ongoing by design. If a service's pitch to you doesn't account for these limits, that's worth noticing. The honest version of this category is "we automate a tedious, repeating chore and save you time" — not "we make you invisible." Take Priya Nair, who signed up for a removal service after a client found her home address through a people-search site. Six months later, her broker listings were mostly gone, which was real progress — but a local news article from years earlier, where she'd been quoted by name in a story about a neighborhood dispute, was still right where it had always been. Her removal service never claimed it could touch that article, and it couldn't have; it simply wasn't the kind of data these services are built to reach. Knowing that going in would have saved her the disappointment of expecting more than the service was ever designed to deliver. Is this a scam? Mostly, no — but the question deserves a direct answer rather than a dismissal. The established names in this space are automating a real, documented, legitimately tedious process. Broker opt-out forms exist, they do work (imperfectly and not permanently, but they work), and a service that submits them on your behalf for a monthly fee is providing a genuine convenience, not a fabricated one. The core mechanism is real. Here's the honest caveat on the other side: doing this yourself, for free, is entirely realistic. It takes longer — figuring out which broker sites list you, finding each opt-out form, and submitting requests across dozens of sites is a real afternoon-or-more project, and it's one most people start and then quietly abandon after the fifth or sixth site because the forms are designed to be a little annoying. If your time is worth more to you than the subscription fee, paying for automation is a reasonable trade. If you have the patience and want to keep the money, the manual path works too — it's the same underlying action, just done by hand. The red flags that do mean scam or overclaim The line between "legitimate automation service" and "overclaiming" shows up in specific promises. Watch for: Guaranteed, complete, permanent erasure from the internet. No service can promise this. The internet has no single control panel, broker sites re-list people, and plenty of information about you exists outside any broker's database entirely. Any absolute claim here should make you skeptical. A promise to erase you from ChatGPT, Gemini, or other AI tools. This is the clearest overclaim to watch for. No company — including the AI providers themselves — can guarantee that a person's data is surgically removed from a trained model on demand. You can submit requests to AI companies regarding your personal data, and those requests are reviewed through their own privacy processes, but "guaranteed removal from AI" is not something any legitimate service can sell you, because it isn't a capability that exists in a controllable, on-demand form. Vague "dark web monitoring" claims with no specifics on what's actually monitored or how. This phrase gets attached to a lot of products loosely; it's worth asking what data source is actually being checked. Pressure-heavy sales pages that emphasize fear over function. A legitimate service can explain plainly what it does and doesn't do. One that leans entirely on alarming hypotheticals about identity theft to sell a subscription is worth reading more skeptically. So should I pay for one, or do it myself? Both are legitimate choices, and the honest answer depends on your own time versus money calculation. If you want this handled without thinking about it again, and you're comfortable with an ongoing subscription, a service like the ones above automates a real process reasonably well. If you'd rather spend a weekend doing it manually and keep the subscription cost, that's equally valid — you're doing the same opt-out submissions a paid service would do, just without paying for the automation and the recurring re-checks. Will any of these stop AI assistants from mentioning me? No, and any service implying otherwise is overstating what's possible. Data broker removal services operate entirely within the world of broker databases and the search results built on them. AI assistants are a separate system with separate data sources, and nothing in a broker opt-out touches what a model says when someone asks it about you directly. That last point is worth acting on regardless of which removal path you choose. Before you spend money on any of these services — or even before you start the manual opt-out project — it's worth taking two minutes to see where you actually stand with AI right now. A free AI name check before you pay anyone shows you what ChatGPT, Gemini, and Claude currently say if someone asks about you by name, which is useful context regardless of what you decide to do about your data broker footprint. And if you haven't yet worked through the broker opt-out process itself, this guide to data broker opt-outs walks through doing it manually, site by site, for anyone who'd rather skip the subscription.
Why AI Confuses You With Someone Else (and How to Stop It)
Ask ChatGPT about David Chen and you might get a software engineer in Seattle. Ask again a week later, or ask Gemini instead, and you might get a restaurant owner in Toronto, or a university researcher who published a paper in 2019. If none of these is actually you, welcome to one of the stranger side effects of how AI assistants work — and one that almost nobody talks about, even though it's increasingly common. Why AI blends you with other people who share your name An AI assistant doesn't have a dedicated file on "David Chen" the way a private investigator might. When it's asked who someone is, it's effectively reconstructing an answer from everything in its training data — and often, anything it can find in a quick web search — that's associated with that exact name string. If there are five different David Chens scattered across the internet with varying amounts of content about them, the model isn't cleanly separating them into five distinct people. It's drawing from a pooled, blended impression of all of them, and it tends to lean hardest on whichever version has the most, or the most prominent, content attached to that name. A few factors make this more or less likely to happen to you: How common your name is. "David Chen" or "Maria Lopez" simply has more people attached to it than "Thessaly Okonkwo-Reyes." More people sharing a name means more raw material for the model to blend together. Whether you share a name with someone more famous or more online. If there's a well-known person, a prolific writer, or just someone with an unusually large digital footprint who happens to share your name, that person's content tends to dominate the pooled answer, even when the question was really about you. How much unique digital footprint you personally have. If you have very little content clearly, consistently tied to your exact name, there's not much for the model to latch onto specifically — which leaves more room for someone else's information to fill the gap. None of this is a flaw unique to one AI company. ChatGPT, Gemini, and Claude all work this way to varying degrees, because they're all solving the same underlying problem: turning a name with no built-in unique identifier into a single, coherent-sounding answer, even when the name itself isn't unique at all. It's also worth saying plainly: this is a genuinely under-discussed corner of how people experience AI. Most advice about "what AI says about you" assumes the problem is something AI got wrong about you specifically. The namesake mix-up is a different animal — the AI isn't wrong about facts so much as it's answering about the wrong person entirely while sounding just as confident as if it had the right one. That's a distinct problem worth naming on its own, not a subset of general AI inaccuracy. Once you understand why this happens, the fix is less about "correcting the AI" directly — you generally can't edit a model's answer the way you'd edit a wiki page — and more about giving it stronger, more specific signal to work with. Before getting into that, though, it's worth being clear on why this is worth fixing at all, beyond simple annoyance. Why a mix-up is more than just annoying It's easy to shrug this off as a quirk — mildly funny the first time, forgettable the second. But a wrong AI answer that's actually describing someone else can genuinely mislead the person reading it, and increasingly, that person is making a decision about you. Picture Maria Lopez, who runs a small accounting practice, being checked out by a prospective client who asks an AI assistant "who is Maria Lopez, accountant" before booking a consultation. If the assistant blends her in with a different Maria Lopez — wrong city, wrong credentials, or worse, a different professional history entirely — that prospective client walks into the call with a false impression, or decides not to call at all based on information that was never about her. She never finds out why the call didn't happen. The AI didn't just get a fact wrong; it quietly handed someone an inaccurate basis for a decision about her. The same dynamic plays out with employers screening a candidate, landlords checking a prospective tenant, or a new acquaintance doing a casual background check before a first date. In every version, the person being mixed up has no idea it happened, and no chance to correct the record in the moment, because the mix-up occurs inside someone else's private conversation with an AI assistant, not in a public place where you could notice and respond. How to give AI a stronger, more specific signal You can't directly edit what a model says about you, and there's no dashboard where you flag "that's not me" and have it instantly fixed. What you can do is make your own identity easier for these systems to isolate from everyone else sharing your name — the same way making a louder, more distinct signal helps you stand out in any noisy room. Use your name the same way, everywhere If your LinkedIn says "David T. Chen," your personal site says "David Chen," and your old Twitter account says "Dave Chen," you've effectively split your own digital footprint into three smaller, weaker signals instead of one strong one. Pick one consistent version of your name — middle initial or not, nickname or full name — and use it the same way across every profile, bio, and byline you control going forward. Consistency doesn't erase the other David Chens, but it makes your own cluster of information denser and easier for a model to recognize as belonging to one person. Build one or two strong anchor profiles Pick a small number of profiles — a LinkedIn page and a personal website are the two that tend to carry the most weight — and make sure each one clearly, prominently states your city and your role: "Maria Lopez, CPA based in Denver," not just "Maria Lopez." Specificity is what gives a model something concrete to grab onto. A profile that just says a name with no distinguishing detail gives the model nothing to disambiguate you from anyone else with the same name; a profile that pairs your name with a specific job and location gives it an actual anchor. Add structured markup if you have a personal site If you maintain your own website, adding Person schema markup — a structured, behind-the-scenes way of telling search engines and AI crawlers "this page is specifically about a person named X, who does Y, located in Z" — gives automated systems an explicit, unambiguous signal rather than asking them to infer it from plain text. It's a technical step, but not a hard one, and it tends to carry outsized weight for exactly this kind of disambiguation problem. We cover how to actually add it in our guide to Person schema markup. Link your profiles together with sameAs connections In plain terms, a "sameAs" link is a way of explicitly telling search engines and AI systems "this profile and that profile are the same person" — for example, having your personal website list your LinkedIn and professional Twitter as sameAs references, and vice versa where the platform allows it. On its own, each individual profile is just one data point. Linked together explicitly, they form a single, connected identity that's much harder for an automated system to confuse with an unrelated stranger who happens to share your name. What's realistic to expect, and when to check again None of these fixes work instantly, and none of them come with a guarantee — you're influencing the signal available to these systems, not flipping a switch inside them. It can take time for updated profiles and new structured data to be picked up, re-crawled, and reflected in how an AI assistant answers. That's a reasonable tradeoff for something that costs you an afternoon of profile cleanup rather than money. The most useful habit is simply checking back periodically rather than doing this once and assuming it worked. Ask the assistants directly every so often, or use a free check that does it for you across several assistants at once — get an alert when AI mixes you up with a namesake so you know the moment the picture changes, instead of finding out by accident the way most people do. Will this fix happen on just one AI assistant, or all of them? Since ChatGPT, Gemini, and Claude pull from different training data and have different web-search habits, a fix that resolves the mix-up on one may not immediately resolve it on another. Stronger, more consistent signal across your own profiles tends to help across all of them over time, but it's worth checking each one rather than assuming a fix on one assistant means the problem is solved everywhere. Is this worth doing if the mix-up isn't hurting me in any obvious way? If you're not job hunting, dating, or running a business under your own name, it's genuinely lower stakes, and you can treat this as optional housekeeping rather than urgent. But the nature of the problem is that you usually don't know when someone is quietly checking you out — which is exactly why the people who benefit most from fixing this are often the ones who didn't realize it was an issue until they checked.
How to Remove Your Personal Information From ChatGPT and Other AI
Someone types their own name into ChatGPT, half out of curiosity, half out of dread, and gets back a paragraph that's wrong, outdated, or something they'd rather not see tied to their name at all. The instinct that follows is immediate: how do I get that removed? It's a fair question, and the honest answer is more complicated than "click delete" — but there is a real, practical path forward, and it's worth walking through calmly rather than panicking or giving up. The part nobody likes hearing: you can't make an AI "forget" a fact on command Large language models like ChatGPT, Gemini, and Claude don't store information the way a database or a social media profile does. There's no row in a table with your name on it that a support agent can find and delete. Instead, these models were trained on enormous amounts of text, and in that process they absorbed general patterns and associations — including, sometimes, facts tied to specific names. Once that training has happened, there is no button anyone can press to surgically pull a single fact back out of the model's underlying weights. This matters because a lot of services — and a lot of worried searches online — imply otherwise. If you see a company promising to "erase you from AI" or "delete your data from ChatGPT's training," treat that claim skeptically. No outside company has the ability to reach into OpenAI's, Google's, or Anthropic's trained models and remove a specific fact about a specific person. That's simply not how the technology works, and anyone claiming they can do it for you is overstating what's possible. That doesn't mean you're stuck. It means the goal needs to be reframed: not "erase the fact from the model" but "change what the model is willing to say back when someone asks about me." That second goal is realistic, and it's the one worth pursuing. What you can actually do OpenAI, Google, and Anthropic each provide some kind of process for raising concerns about personal data appearing in their products, including requests related to privacy or inaccurate information about a named individual. The specifics of each company's form, the information they ask for, and how they handle the request differ, and they change over time, so it's worth checking each company's current help or privacy pages directly rather than relying on instructions from a blog post. In general terms, you'll typically need to identify yourself, describe what the AI is saying, and explain why you believe it's inaccurate or shouldn't be surfaced. Consider a fictional example: David Chen notices that when people ask ChatGPT about him by name, it repeats an outdated claim connecting him to a company he left years ago, worded in a way that sounds like he still works there. David can't reach into the model and delete that association. But he can submit a request through the appropriate channel describing exactly what's wrong and why, in the hope that it gets corrected in future answers. It's worth doing this for each AI provider separately. A request to one company has no effect on how a different company's model answers, because they're trained and operated independently. If something false is showing up across ChatGPT, Gemini, and Claude, that's three separate requests, not one. What a request can realistically achieve Here's the distinction that makes all of this make sense: the training itself generally can't be undone, but what the model is willing to say back to a user can change. AI companies have ways to adjust, filter, or restrict a model's output after training — sometimes in response to a specific complaint about a specific person or claim. So even though the underlying training data isn't being edited line by line, the practical result a person experiences can still shift: the model may stop repeating a specific claim, decline to answer, or give a more cautious, hedged response instead of stating something as fact. This is genuinely useful, even though it sounds less dramatic than "deletion." If the thing you care about is what a stranger sees when they ask an AI assistant about you, a change in the model's output is the thing that actually matters to that stranger — not what's technically still sitting somewhere in a training set they'll never look at directly. It's also worth being realistic about limits. Not every request results in a visible change. Some claims are broad patterns learned from many sources rather than a single traceable fact, and those are harder to isolate and adjust. Some requests are reviewed and declined. There's no guaranteed outcome here, and no legitimate service — including ours — can promise one. Timing, legal rights, and what this isn't These requests are not instant. Expect a review process that can take time, and don't assume silence means rejection — or that a reply means success. Checking back periodically is the only reliable way to know whether anything changed. If your situation involves a legal right to have data corrected or removed — for example, under GDPR in the EU or similar consumer privacy laws like the CCPA in California — that's a different and potentially stronger lever than a general feedback form, and it's a path worth understanding on its own terms. This piece on GDPR and CCPA deletion requests covers how those formal rights work. One important note: nothing in this article is legal advice. If your situation involves defamation, harassment, or a genuine legal claim, talk to a lawyer who handles privacy or online reputation issues — the stakes are higher than a blog post can responsibly address. Does asking an AI company to remove information cost anything? No reputable AI provider charges a fee to submit a data or privacy-related request. If a third party asks you to pay them specifically to "submit a deletion request to ChatGPT" on your behalf, be skeptical — you can typically submit these requests yourself directly with the AI company at no cost. Will deleting my old social media accounts make the AI forget me too? Not directly, and not quickly. Deleting a source page doesn't retroactively edit a model that was already trained on text that existed at some point in the past. It may help reduce how much new information about you is available for any future model training, but it won't change what an already-trained model currently says. How do I even know what an AI is currently saying about me? Ask it directly, phrased a few different ways, across more than one AI assistant, since each one answers independently. That's also the exact gap Tailwind's free personal reputation check was built to close — it's a quick way to see whether AI still repeats it across ChatGPT, Gemini, and Claude in one place, without you having to interrogate each one by hand. None of this is as satisfying as a clean "delete" button. But submitting the requests that genuinely apply to your situation, understanding what they can and can't do, and then checking back later to see whether the answer has actually changed, is a real strategy — just not an instant one.
How to Check Your Digital Footprint in 30 Minutes
Most people have a rough, outdated mental picture of what's findable about them online — a guess based on what they remember posting years ago, not what's actually visible today. The only way to close that gap is to actually look. This is a timed checklist you can run through in about 30 minutes, covering search engines, old social accounts, data broker sites, breach exposure, and — increasingly relevant — what AI assistants say when someone asks them about you directly. Before you start Grab a notes app or a blank document. You're not fixing anything yet — you're just collecting a snapshot of what's out there right now, so you have something to compare against next time. That last step matters as much as the search itself, and we'll come back to it at the end. One quick preview: the last step on this list is asking AI assistants what they say about you by name, and you can run that one in a single pass with a free tool that checks ChatGPT, Gemini, and Claude at once — add AI answers to your footprint check alongside the more familiar search and social steps below. 1. Search engines (about 5 minutes) Search your full name in quotes — "Jane Martinez" rather than Jane Martinez — to cut down on irrelevant results with the same first and last name. Do this on at least two search engines, since results genuinely differ. Then repeat the search under the Images and News tabs specifically. People often forget that image search surfaces old photos tagged with their name years after they'd otherwise have forgotten about them, and news search can surface local news mentions, old press releases, or court/public records that don't show up in a standard web search. 2. Social media accounts (about 5 minutes) List every social platform you've ever had an account on, including ones you stopped using years ago — old Twitter/X, an abandoned LinkedIn, a Facebook profile from college, a Myspace account if you're old enough for that to be real. For each one, check three things: is it still public, is the information on it still accurate, and would a stranger searching your name actually find it. An old, abandoned, but still-public profile with an outdated job title or a photo from a decade ago is a common and easy-to-fix gap. 3. Data broker and people-search sites (about 5 minutes) This isn't the moment to start a full opt-out project — that's a bigger undertaking on its own. Right now, just spot-check two or three well-known people-search sites by searching your name and seeing what comes up: your current or past home addresses, phone number, relatives' names, and age are the typical fields exposed. The point of this step is visibility, not cleanup — knowing whether your information shows up at all shapes whether opting out is worth prioritizing later. 4. Breach exposure (about 5 minutes) Check whether your email address has shown up in a known data breach using a reputable, free breach-check tool like Have I Been Pwned. It's a long-standing, free public service, not a commercial product trying to sell you something, and it will tell you which known breaches your email has appeared in and roughly what kind of data was exposed in each one — passwords, phone numbers, or more. If you find your email in multiple breaches you don't recognize responding to, that's a signal to update passwords on those specific accounts and turn on two-factor authentication where you haven't already. 5. AI assistants (about 5 minutes) This is the newest piece of a digital footprint check, and the one most people haven't thought to run yet. Open ChatGPT, Gemini, and Claude separately and ask each one directly: "What do you know about [your full name]?" or "Who is [your full name]?" Try a couple of phrasings, since these systems can give noticeably different answers depending on how a question is worded, and each AI assistant answers completely independently of the others — a clean answer from one doesn't tell you anything about what another one says. If that sounds like a lot of separate tabs and typing, a faster route is running a single check across all three at once — that's exactly what the free tool at Tailwind's personal reputation checker does, producing one page you can scan rather than three separate conversations to compare by hand. 6. Log what you found (about 5 minutes) Go back to that notes document and write down, in plain bullet points, what you found at each step: which old accounts are still public, whether any data broker sites had your info, whether your email turned up in a breach, and what each AI assistant said about you. This isn't about fixing everything today. It's a baseline — a record of where things stood on this date, so the next time you run this same checklist, you can tell whether things got better, worse, or stayed the same, instead of starting from scratch with no memory of what you already checked. How often should I actually repeat this? Every few months is a reasonable rhythm for most people — often enough to catch new exposure before it sits unnoticed for years, infrequent enough that it doesn't become a chore. If you've recently changed jobs, moved, gone through something public (a lawsuit, a news mention, a layoff announcement), or noticed something specific that worried you, that's worth an off-cycle check rather than waiting for the next scheduled one. Do I need to fix everything I find in one sitting? No — and trying to will probably make the whole exercise feel overwhelming enough that you avoid doing it again. Treat this checklist as reconnaissance, not a to-do list that has to be cleared the same day. Note what you found, then prioritize: an inaccurate claim repeated by an AI assistant or a data broker listing your home address is usually worth addressing sooner than an old, harmless social post. Thirty minutes once every few months is a small, sustainable habit compared to the alternative — finding out what's publicly tied to your name only when someone else brings it up first. For a deeper dive into the search-engine piece specifically, this guide on how to Google yourself properly goes further into search operators and techniques that go beyond a basic name search.
Person Schema Markup: Tell Google and AI Exactly Who You Are
Search your own name and you might land on a dentist in another state, a defunct musician, or someone who shares your name and got arrested for something you've never done. Search engines and AI assistants don't know who you are the way a friend does. They're reading unstructured text scattered across the web and making a best guess about which facts belong to which person. Sometimes that guess is wrong, and your name ends up tangled up with somebody else's biography, photo, or bad press. Why Search Engines and AI Have to Guess When you write "Maria Lopez, Pastry Chef at Sweetleaf Bakery" on your own website, a human reader understands instantly who that sentence is about. A crawler reading the same page has to work harder. It sees a string of text, not a verified fact. It doesn't automatically know that the "Maria Lopez" on your About page is the same "Maria Lopez" who has a LinkedIn profile, an Instagram account, and a mention in a local newspaper. If there's another Maria Lopez online — and there almost certainly is — the crawler can mix up details between the two of you, or an AI model summarizing "who is Maria Lopez" can blend facts from both into one confused answer. This is the core problem behind a lot of mistaken-identity situations online: there's no explicit label saying "this specific person is the subject of this specific page." Structured data exists to fix exactly that gap. Instead of hoping a crawler correctly infers facts from paragraphs, you hand it the facts directly, in a format built for machines rather than humans. The stakes of this guessing game are higher than they used to be. It isn't only about where a page ranks in search results anymore. AI chatbots now routinely get asked questions like "who is Maria Lopez" or "tell me about [your name]," and they answer by pulling together whatever text they've encountered about anyone who shares that name. If the available information is thin or ambiguous about who you specifically are, the model fills in the gaps with the nearest match it has — which might be a stranger. Structured data doesn't rewrite an AI model on command, but it does make your own corner of the internet unambiguous, and that matters more as online reputation gets filtered through AI summaries instead of a page of search results. Person Schema and the sameAs Property Schema markup is a shared vocabulary that search engines and many AI crawlers understand. One piece of that vocabulary is the Person type, which lets you explicitly state things like a person's name, job title, location, and employer, directly in the code of a webpage. It's invisible to visitors — it doesn't change how the page looks — but it gives crawlers a clean, structured fact sheet instead of a paragraph to interpret on their own. The most useful property for anyone dealing with online confusion is sameAs. In plain terms, sameAs is how you tell a crawler "this LinkedIn profile, this Twitter/X account, and this website are all the same person." You list the URLs of your other profiles, and that list becomes a signal tying them together as one identity. Instead of a search engine treating your personal site, your LinkedIn page, and your Instagram as three separate, unrelated things it has to guess are connected, you've told it directly that they belong to the same person. This matters a lot when your name isn't unique. If there are five Maria Lopezes with public profiles, sameAs is one of the clearest ways to draw a box around which accounts belong to you specifically, rather than leaving the connection to inference. It's also one of the few signals you control completely — you're not asking anyone's permission to state which profiles are yours, you're simply publishing that fact in a format machines can read. A Working Example You Can Copy Below is a complete, working example for a fictional person — Maria Lopez, a pastry chef with a personal website, a LinkedIn profile, and an Instagram account. This is written in JSON-LD, which is the format search engines and most AI crawlers prefer for structured data. You can adapt it by swapping in your own name, job title, and profile URLs. <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "Person", "name": "Maria Lopez", "jobTitle": "Pastry Chef", "url": "https://marialopezbakes.com", "image": "https://marialopezbakes.com/images/maria-lopez.jpg", "worksFor": { "@type": "Organization", "name": "Sweetleaf Bakery" }, "sameAs": [ "https://www.linkedin.com/in/marialopezbakes", "https://www.instagram.com/marialopezbakes", "https://twitter.com/marialopezbakes" ] } </script> A few notes on what's happening here. @type: "Person" tells the crawler this block of data describes a human being, not a business or a product. name and jobTitle are the basic facts. url points to the page this data lives on, which anchors everything else to that specific website. And sameAs is the list doing the heavy lifting — each URL is another place on the internet where this same Maria Lopez has a presence, so a crawler piecing together "who is Maria Lopez" has a direct, machine-readable trail to follow instead of a guess to make. You don't need every property filled in. If you don't have an employer to list, leave worksFor out entirely rather than putting in a placeholder. The same goes for any field that doesn't apply to you — an empty or made-up value is worse than no value at all, since it adds noise instead of clarity. Where to Actually Put This Code If you manage your own website's code directly, this script block goes in the <head> section of the page that's about you — usually your homepage or an About page. It doesn't need to go on every page of the site, just the one meant to represent you as a person. If your site runs on WordPress, Squarespace, Wix, or a similar platform, you generally won't be editing raw HTML files directly. Most of these platforms have either a built-in field for custom code in the page header, or a plugin that lets you add JSON-LD structured data without touching the rest of your site's code. The exact menu location varies by platform and changes over time, so it's worth checking your platform's current settings or plugin directory rather than following outdated instructions. Once it's added, you can use a structured data testing tool to confirm the markup is valid and being read correctly, rather than just trusting that you copied it right. These tools will flag typos, missing commas, or fields that don't match what the schema expects, which is easy to get wrong when copying by hand. Do I need to know how to code for this to work? Not really. You're not writing new code from scratch — you're copying a template, swapping in your own details, and pasting it into one field. The hardest part for most people is finding where their website platform lets them add custom header code, not writing the JSON-LD itself. Will this immediately stop AI from confusing me with someone else? No single fix works instantly, and schema markup doesn't come with any kind of guarantee — it's a signal, not a switch. AI models are trained on huge amounts of existing text, and that training data isn't rewritten the moment you add a script tag to your site. What schema markup does is make your own, current, correct information easier for new crawls and future lookups to parse and trust over time. It's one piece of a larger effort, not a one-step fix, and it tends to help more the longer it's been in place. What if I have more than one website or profile that represents me? Add the same Person schema, with the same sameAs list, to each site you control. Consistency is the point — if your personal site, your professional portfolio, and a side project all point to the same set of profiles as "this is the same person," that repetition reinforces the signal rather than diluting it. One Piece of a Bigger Picture Person schema markup won't rewrite anything that's already out there about you, and it won't untangle every mix-up instantly. But it's one of the few concrete, doable things you can set up once and leave running — a clear, standing signal about who you actually are, sitting on a page you control. If you're dealing with the broader problem of being confused with someone else online, it's worth reading why AI sometimes confuses you with someone else for the full picture of how these mix-ups happen and what else helps. If you run a local business under your own name — a salon, a private practice, a studio — a similar structured data approach applies to the business itself. Our LocalBusiness schema generator can help you build that version without writing any code. And once you've made a change like this, the natural next step is checking whether it's actually landing. Tailwind's free AI name check lets you see what ChatGPT, Gemini, and Claude currently say when asked about you, so you can keep AI from linking you to the wrong person and track whether your profile is coming through clearly as more of it builds up over time.
How to Get a Google Knowledge Panel for Yourself
You've probably seen a Knowledge Panel before without thinking much about it — search for a well-known actor, a local mayor, or a famous chef, and a box appears on the right side of Google with their photo, a short bio, and a handful of facts. Then you search your own name and get nothing like that, just a list of blue links. It's natural to wonder if there's a button somewhere to get one of your own. There isn't, and understanding why is more useful than any shortcut would be. What a Knowledge Panel Actually Is A Knowledge Panel is the summary box Google shows for an entity it considers well-established — a person, a place, a company, a band, a landmark. It pulls together facts like an occupation, birthplace, notable works, or official website, sourced from what Google calls its Knowledge Graph: a massive, continuously updated database of entities and the verified facts connecting them. The panel isn't written by a human editor sitting at Google reviewing your case. It's generated automatically when Google's systems decide there's enough notable, verifiable, cross-referenced information about an entity to summarize with confidence. That last part is the key thing to understand before chasing one. A Knowledge Panel isn't something you apply for, and it isn't something you can buy, request, or submit a form to receive. It shows up as a side effect of already having an established public presence — generally some mix of press coverage, a Wikipedia page, consistent and verified profiles across major platforms, and structured data that makes your identity unambiguous across the web. If that presence doesn't exist yet, no amount of asking will manufacture a panel out of nothing. It's worth being direct about this up front: most people, including plenty of successful, accomplished people, will never have a Knowledge Panel, and that's not a sign of failure. It's a reflection of how Google's systems are built, not a verdict on anyone's worth or achievements. It also helps to understand what the panel is pulling from. Google's Knowledge Graph isn't built from any single source — it cross-references facts that show up repeatedly and consistently across many independent places: news archives, Wikipedia and Wikidata, verified business and professional listings, government and institutional records, and structured data published by websites themselves. When the same fact about you — your occupation, your employer, your location — shows up the same way across enough independent sources, Google's confidence in that fact increases. When sources conflict, or when there isn't enough independent material to cross-reference in the first place, there's nothing solid enough to summarize into a panel. This is also why two people with an identical name can end up in very different positions: one might have enough disambiguating, verifiable material online for Google to construct a confident picture, while the other doesn't, regardless of how deserving either person might feel they are. The Realistic Path Toward Becoming Eligible There's no official checklist Google publishes that guarantees a panel if you complete every item. But based on what's observable across people and organizations who do have one, a few contributing factors show up consistently. A well-maintained, verified presence across major platforms. LinkedIn, professional directories, publisher bylines, official social accounts — kept current, consistent, and clearly attributed to you specifically, not a stale profile from years ago. Consistent structured data. Google's systems lean heavily on structured signals to confirm that different mentions of a name across different sites refer to the same entity. If you haven't set this up on your own site yet, adding Person schema markup is a concrete way to make that identity explicit rather than something Google has to infer. Genuine media coverage or notable achievements. Independent, third-party coverage — news articles, interviews, industry publications writing about you because you did something notable — carries far more weight than anything you publish about yourself. This is usually the hardest factor to manufacture, because it generally has to be earned rather than created on demand. A Wikipedia or Wikidata entry, if you're genuinely notable enough for one. This deserves an honest caveat: Wikipedia has strict, actively enforced notability standards, and most people, including many who are quite successful in their field, don't meet them. Attempting to create a Wikipedia page about yourself, or paying someone to do so when you don't clearly meet those standards, tends to backfire — undisclosed self-promotional articles get flagged and removed, and editors are specifically on the lookout for this pattern. If you're genuinely notable by Wikipedia's standards, a page can happen through the normal editorial process. If you're not there yet, trying to force it isn't a shortcut worth taking. None of these factors is a guarantee on its own, and there's no confirmed formula combining them that produces a panel on a predictable timeline. They're contributing conditions, not a checklist with a finish line. A few other things that tend to help on the margins, without being decisive on their own: an official, verified website that's clearly about you and kept up to date; consistent use of your full name rather than switching between variations across different platforms; and having your occupation or affiliation confirmed by an organization other than yourself, such as an employer's staff page or a professional association's member directory. None of these will single-handedly produce a panel, but they add to the body of consistent, cross-referenced material that makes one more plausible over time. If You Do Get One, How to Claim or Manage It If a Knowledge Panel does eventually appear for you, Google provides some mechanism for a verified individual to suggest edits or claim partial ownership of their own panel's content — generally involving confirming your identity through a connected account and then being able to flag incorrect facts or suggest updates. The specific steps, verification requirements, and what you're able to edit directly versus merely suggest have changed over time, so it's worth checking Google's current help documentation rather than relying on older instructions you might find elsewhere. Even with a claimed panel, you're generally not able to rewrite it freely the way you'd edit your own website. It still reflects what Google's systems believe is accurate based on available sources, and factual corrections typically need some kind of supporting source behind them rather than just your say-so. If you spot something outright wrong in a panel — a mismatched photo, a job title that isn't yours, a birthdate that's off — there's also generally a feedback mechanism for flagging the error even if you haven't formally claimed the panel. This tends to be a slower, less direct process than claiming the panel as a verified owner, and it doesn't guarantee a correction, but it's worth using if you notice a factual mistake rather than leaving it uncorrected. One more thing worth flagging: having a panel doesn't freeze its contents. If your public information changes — a new job, a new location, a notable project — the panel can update on its own as Google's systems reprocess the web, sometimes faster than you'd expect and sometimes slower. Claiming the panel gives you a more direct way to nudge corrections, but it doesn't put you in full editorial control of what appears. Can I just pay someone to get me a Knowledge Panel? Be skeptical of anyone promising this as a guaranteed, paid service. Reputable SEO and PR work can help build the underlying conditions — media coverage, structured data, a consistent cross-platform presence — that make a panel more plausible over time. But nobody can sell you the panel itself, because it's generated algorithmically based on Google's own assessment of your notability, not issued by a sales team. If an offer sounds like it's selling the panel directly rather than the groundwork underneath it, treat that as a red flag. How long does it take to get a Knowledge Panel once I meet the criteria? There's no fixed timeline, and Google doesn't publish one. It depends on how quickly its systems crawl and reprocess the relevant signals, which can be a matter of weeks or considerably longer. This is part of why chasing a panel as a short-term goal tends to be frustrating — it's a lagging indicator of an established presence, not a fast result from a specific action. Why Most People Shouldn't Make This the Goal Here's the honest bottom line: a Knowledge Panel is a reasonable aspiration if you're a public figure, a journalist, an author, a founder with real press coverage, or otherwise building a genuinely notable public profile. For the vast majority of people — a small business owner, a freelancer, someone simply trying to manage how they appear online — it's not a realistic goal, and treating it as one is likely to lead to wasted effort chasing something that isn't coming. What's actually useful for almost everyone, regardless of whether a Knowledge Panel is ever in the cards, is making sure your basic information is accurate wherever it does show up — including in AI assistants, which increasingly answer questions about people the way a Knowledge Panel summarizes them, but without the same sourcing standards or editorial oversight. The things that help with a panel — structured data, a consistent presence, accurate public information — also directly help AI assistants describe you correctly, and unlike a Knowledge Panel, you don't need to wait for anyone's algorithm to decide you're notable enough to check that. There's also a meaningful difference in how forgiving each system is of an ordinary, non-famous life. Google's Knowledge Graph is explicitly built around notability — it's trying to model entities the public is likely to search for, which inherently excludes most people most of the time. An AI assistant answering a direct question about a specific name doesn't have that same bar. Ask it who David Chen is, and it will try to answer based on whatever it can find, whether David Chen is a famous architect or a private individual with a small professional footprint. That makes checking what AI says about you a far more universally applicable exercise than chasing a panel that, for most people, simply isn't coming. Start with the structural piece — our guide to adding Person schema markup to your own site covers the mechanics. Then, as your presence grows or changes, Tailwind's free AI name check lets you see exactly what ChatGPT, Gemini, and Claude currently say about you, so you can track how AI describes you as your profile grows instead of waiting to find out by accident. There's a practical advantage to checking AI assistants that a Knowledge Panel doesn't offer: you get an answer today, for free, regardless of how notable you are. You don't need press coverage or a Wikipedia entry to ask ChatGPT, Gemini, or Claude what they say when someone types your name, and the answer tells you something real — whether the facts are accurate, whether you're being confused with someone else who shares your name, or whether there simply isn't enough reliable information out there yet for the model to say much of anything. That's useful information either way, and it's available to literally everyone, which a Knowledge Panel will never be.
Data Broker Opt-Out List: How to Remove Yourself From People-Search Sites
A list like this one is a snapshot, not a permanent reference — people-search sites change their opt-out processes, rename their forms, and occasionally get acquired or merged into a different company, so the steps that worked last year might look different today, and it's worth checking a site's own current opt-out page rather than assuming nothing has moved since this was written. What Data Broker and People-Search Sites Actually Are Data brokers are companies that collect personal information from public records, marketing databases, social media, and other brokers, then package it into searchable profiles they sell access to — sometimes to businesses doing background checks, and sometimes directly to anyone willing to pay a few dollars for a report. People-search sites are the consumer-facing version of this: you type in a name, and the site surfaces an address history, phone numbers, relatives' names, age, and sometimes a lot more, usually behind a paywall that charges you to see the "full report." These sites matter more than people often realize because of where they tend to rank. Search someone's name, and a people-search listing frequently shows up on the first page of results — sometimes higher than anything the person actually published themselves. For someone dealing with an ex, a stalker, a scammer impersonating them, or just basic privacy concerns, having an address and phone number sitting a few clicks away from their name is a real exposure, not an abstract one. It's also worth noting that these listings aren't always accurate, even though they're often treated as authoritative. Ages get miscalculated, old addresses get listed as current, and "possible relatives" sections sometimes link you to people you've never met who simply shared an address or a last name at some point in a public record. Part of the value of checking these sites isn't just about privacy exposure — it's also about catching information that's simply wrong and sitting next to your name regardless. The data itself usually isn't stolen or hacked — it's compiled from sources that are technically public, like property records, court records, and voter registration files, then cross-referenced and resold at scale. That's part of why opting out of one site doesn't solve the underlying problem: the same underlying data keeps getting re-licensed and re-scraped into new listings elsewhere. It's worth understanding roughly how these sites build a profile, because it explains why the same information keeps showing up across so many of them. A broker typically starts with a core set of public records — property deeds, court filings, voter rolls — then layers in data purchased from marketing firms, app permissions, loyalty programs, and other brokers who've already aggregated their own slice of the picture. The result is a profile that's often more complete than any single public record on its own, which is exactly why these listings tend to feel more invasive than, say, a standalone county property record would. Because many brokers buy and sell data between each other, a single opt-out rarely touches the whole web of sources — it mainly removes the specific site's own public-facing listing. The Sites Worth Checking First These are among the most visible and widely used people-search platforms. Search your own name on each one before assuming you're not listed — many people are surprised to find a listing on a site they've never heard of. Spokeo — one of the largest and longest-running people-search sites, aggregating contact info, social media links, photos, and public records into a single profile. It's often one of the first results for a name search, which makes it a reasonable starting point for checking what's out there about you. WhitePages — originally an online phone directory, now a broader people-search service covering addresses, phone numbers, and background information. It also runs a separate paid identity-protection product, which is distinct from its free-to-browse people-search listings. BeenVerified — markets itself primarily as a background-check tool, but functions much like the others: search a name, get a preview showing an age range and possible city, pay for the full report to see more. Intelius — another long-standing background-check and people-search service, often appearing in search results alongside its sister sites, since several of these companies share ownership or data infrastructure. PeopleFinders — focused specifically on locating current and past addresses, phone numbers, and relatives for a given name, with a report format similar to the background-check sites on this list. TruthFinder — a background-check platform that compiles public records, including some criminal and court record data alongside standard contact information, and tends to lean on that record-based framing more heavily than pure directory-style sites. MyLife — notable for generating a "reputation score" alongside contact details, which has drawn criticism for how it presents unverified information and anonymous reviews as if they were a scored, authoritative assessment of someone's character. Radaris — aggregates public records and contact data, and is frequently among the first results when searching a name alongside the more established players, including possible relatives and past addresses going back years. Instant Checkmate — similar in structure to BeenVerified and TruthFinder, combining background-check framing with the same underlying contact and public-record data, often cross-listing the same underlying profile under a different brand. PeopleLooker — a newer entrant in the same space, pulling from similar underlying data sources as the rest of this list and following a similar preview-then-paywall structure. This isn't an exhaustive list — there are smaller, lesser-known people-search sites too, and new ones appear periodically. But these are consistently among the highest-traffic and highest-ranking ones, which makes them the highest-value places to start if you're triaging rather than trying to chase down every possible listing on day one. Most of these sites provide some kind of "remove my info" or "opt out" link, usually in their footer or privacy policy page, and most require you to first locate your own listing on their site before you can submit a removal request for it specifically. The exact process, required steps, and how long removal takes varies by site and changes periodically, so treat the general description here as a starting point rather than a precise, current walkthrough for any one of them. It's also worth knowing that several of these companies are related to each other — some share a parent company or data pipeline even though they operate under different brand names. That doesn't change what you need to do on each individual site, but it explains why removing yourself from one doesn't automatically remove you from what looks like an entirely separate competitor. Before You Start Submitting Requests A little preparation makes the actual opt-out process faster and less confusing once you're clicking through ten different sites in a row. Search variations of your name. A listing might be filed under a maiden name, a middle name you don't normally use, a nickname, or a misspelling. Searching only the exact way you usually write your name can mean missing a listing that's sitting under a slightly different version of it. Note every city or state you've lived in. Many of these sites organize listings by location, and if you've moved, you may have a separate listing tied to a previous address that doesn't show up when you search assuming you'll find yourself under your current city. Check for listings tied to relatives. Because these sites build out family and associate networks, you may find your own information surfacing on a listing built around a parent, sibling, or former roommate, rather than one built around your own name directly. Keep a simple running list. A basic spreadsheet or notes file with the site name, the date you submitted the opt-out, and whether you've confirmed removal gives you something to check back against later, rather than trying to remember which of ten sites you've already handled. Habits That Make This Less Painful Working through even eight or ten of these sites is tedious by design — the process isn't technically hard, it's just repetitive and full of small friction points. A few habits make it go more smoothly. Use a dedicated email address for opt-out confirmations. Setting up a separate inbox just for this process keeps confirmation emails, follow-up requests, and any further correspondence in one place, rather than mixed in with your regular inbox where they're easy to lose track of. Expect to click a confirmation link. Many of these sites don't process a removal the moment you submit the form — they send a confirmation email first, and the request doesn't go through until you click it. Skipping that step is one of the most common reasons an opt-out silently fails. Screenshot or save your listing before you remove it. If a removal doesn't take effect, or you need to follow up, having a record of exactly what was listed and the date you requested removal makes any later back-and-forth easier. Expect listings to come back. This is less a one-time project than ongoing maintenance. Because these sites pull from the same underlying public-record pipelines on a recurring basis, a listing you removed six months ago can simply reappear once the source data gets re-scraped. Treat a first pass through this list as a baseline, not a finish line, and plan to check back periodically. Don't pay for the "report" to see what it says about you. Most of these sites show enough in the free preview to confirm you're listed. Paying for the full report mainly benefits the data broker, not you, and isn't a required step before requesting removal. Watch for opt-out pages that try to upsell you first. Some sites route you through a page promoting their own paid monitoring or removal product before showing the free opt-out option, or bury the free link below the fold. The free opt-out usually still exists even when a paid alternative is presented more prominently. Set a calendar reminder to recheck in a few months. Since listings can reappear, picking a fixed interval — every three to six months, for instance — to re-search your name across this list turns an open-ended worry into a scheduled, bounded task. Is it actually worth doing this myself, or should I just pay someone? That depends mostly on your time and tolerance for repetitive tasks versus your budget. Doing it yourself costs nothing but time, and the process for any individual site usually takes only a few minutes once you've found your listing. The tedious part is the sheer number of sites and the fact that it's not a one-time job. This is exactly the gap that paid removal services like DeleteMe, Incogni, and Aura are built to fill — they automate the searching and submitting across a long list of brokers and handle the recurring re-checks on an ongoing basis so you don't have to remember to come back to it every few months. Whether that's worth paying for depends on how much your time is worth to you, how many sites are actually listing you, and how much peace of mind you get from not having to think about it — we break that comparison down in more detail in whether data removal services are worth paying for. There's also a middle path worth mentioning: you don't have to pick one approach forever. Plenty of people work through this list themselves once to get a feel for what's out there and how invasive their own listings are, then decide afterward whether the ongoing maintenance is worth handing off to a paid service. Doing one manual pass first, even if you eventually pay someone else for the recurring work, at least gives you a clear sense of what's actually being removed on your behalf. Will opting out of these sites stop someone from finding my information another way? Not entirely, and it's worth being realistic about that. People-search sites are one visible, searchable source of this kind of information, but they're compiling data that often exists elsewhere too — public records offices, old social media posts, forum comments, cached pages. Opting out removes a convenient, aggregated, easily searchable version of your information, which meaningfully reduces casual exposure. It doesn't erase the underlying public records these sites draw from, and it's not something that can be done with a single guaranteed action — it's an ongoing reduction in visibility, not a permanent deletion of everything that exists about you. If you're dealing with a situation where exposure carries real safety risk — escaping domestic violence, stalking, a threatening ex, or similar circumstances — treat this list as a starting point rather than a complete safety plan. Specialized privacy and safety resources exist for exactly these situations and generally go further than a standard opt-out routine, covering things like address confidentiality programs and more aggressive takedown options that a general consumer guide like this one isn't equipped to walk through. After You've Worked Through the List Once you've gone through opting out of the major sites, it's reasonable to want some sense of whether it actually changed anything. Search engines can be slow to reflect removals, and the sites themselves rarely tell you definitively that a listing is gone for good. One practical way to check is to see what AI assistants currently say when asked about you — since many of them pull from recent web data and public information in ways that can surface the same kind of address, contact, and background details these broker sites compile. Tailwind's free AI name check lets you see exactly what ChatGPT, Gemini, and Claude say about your name right now, so you can check your name in AI after opting out and get a sense of whether the picture has actually shifted, rather than just hoping it has.
GDPR and CCPA Deletion Requests: Templates That Work
If a company has personal information about you somewhere online and you want it gone, you're not just hoping for a favor. In many cases, you have an actual legal right to ask for it — and the company has an obligation to respond, not just a polite option to ignore you. Two laws come up most often: the GDPR in Europe and the UK, and the CCPA in California. People reach for these requests for all kinds of reasons — an old account they forgot they had, a data broker that scraped their information without asking, a service they stopped using years ago that still has their details on file. Here's what these laws actually give you, in plain terms, plus a template you can copy, paste, and send today. What GDPR's "right to erasure" and CCPA's "right to delete" actually cover Both laws work on a similar idea: if a company holds personal data about you, you can generally ask it to delete that data, and the company generally has to comply unless it has a legitimate reason not to. GDPR (General Data Protection Regulation) applies broadly to residents of the EU and UK, and to companies that process their data, regardless of where that company is headquartered. It includes what's commonly called the "right to erasure" — sometimes called the "right to be forgotten." In plain terms, it lets you ask a company to delete personal data it holds about you. CCPA (California Consumer Privacy Act), and its update the CPRA, gives a similar "right to delete" to California residents. It works on the same basic principle: you can ask a business to delete personal information it has collected about you. Neither right is unlimited. Companies can legally refuse or partially refuse a deletion request for reasons that generally include: Data they're legally required to retain — tax records, billing history, or data needed for regulatory compliance. Data necessary to complete a transaction or provide a service you're still actively using. Data needed to detect fraud or security incidents. Internal uses that are compatible with why you gave the data in the first place, in some circumstances. So a deletion request isn't always an all-or-nothing switch. A company might delete your marketing profile while keeping a record tied to an active account or a past transaction it's required to retain for a set period. Imagine someone like David Chen, who closed an online shopping account three years ago but recently noticed the company still emails him occasional promotions. A deletion request wouldn't necessarily erase his old order history if the company has a legal obligation to keep transaction records for a certain number of years — but it should stop the marketing emails and remove him from any active profile the company is using to target ads. It's also worth understanding that these rights generally apply per company, not as one universal request that clears your name from the internet. If your information is sitting with five different data brokers, you typically need to send five separate requests, because each one is a separate legal relationship with its own obligations. These rights are also not limited to obviously "creepy" uses of your data. Priya Nair, for example, might want to send a deletion request simply because she's switching careers and no longer wants an old employer's HR vendor holding onto her records, not because anything went wrong — the law doesn't require a dramatic reason. None of this is exotic or reserved for lawyers. These requests get filed by ordinary people every day, and most companies — especially larger ones that operate across the EU, UK, or California — have built standard processes to handle them, precisely because the volume of requests is high enough that ignoring them isn't realistic. A quick but important note: this is a simplified, general explanation for people who've never filed one of these requests before — not legal advice. Whether a specific law applies to you, which exceptions a company can legitimately claim, and how to handle a dispute all depend on your jurisdiction and your specific situation. If you're dealing with something high-stakes — harassment, identity theft, a company that's flatly ignoring the law — talk to a lawyer rather than relying on a blog post. A deletion request template you can actually use Below is a template modeled on the kind of language that's commonly used in real deletion requests. Fill in the brackets, pick the law that applies to you (or mention both if you're not sure), and send it. Subject: Data Deletion Request — [Your Full Name] To [Company Name] Privacy Team, My name is [Your Full Name], and my account or customer record may be associated with [Your Email/Account ID]. I am writing to request that you delete the personal data you hold about me, under my right to erasure under GDPR / my right to delete under CCPA (please specify whichever applies to your situation). Specifically, I am requesting that you: 1. Delete all personal data associated with my name, email, and any linked accounts or profiles. 2. Confirm in writing once this deletion has been completed, including the date it was carried out. 3. If you are unable to fully comply with this request, please identify which specific exception you are relying on, and which categories of data are affected. If you need to verify my identity before processing this request, please let me know what information you require. I look forward to your confirmation. Sincerely, [Your Full Name] [Date] A few notes on using this template: swap in the specific law based on where you live, not where the company is based. If you're not sure which one applies, it's fine to mention both and let the company sort out which framework governs your request. If the data in question is something an AI company may have used to train a model rather than just a stored account record, say so specifically — that's different enough from an ordinary customer database that it's worth calling out by name. How to actually send it, and what to expect A few practical habits make these requests much more effective than just firing off an email and hoping. Use the official channel, not general support. Most companies, even small ones, have a dedicated privacy or data-request process — often listed in their privacy policy under a heading like "Your Rights" or "Data Requests." Sending your request there, rather than to a general support inbox, means it reaches the team that's actually responsible for processing it, and it's harder for the company to claim later that it never received it. Keep a copy and a timestamp. Save a copy of exactly what you sent and when. If the company disputes later that it received your request, or if you need to escalate, having a clear record of the date matters. Expect a response window. Both laws generally give companies a defined period to respond — commonly discussed as around 30 days, though the exact statutory window, and the circumstances under which it can be extended, is a legal detail that varies and is worth confirming independently if it matters for your situation. If you're ignored, don't just give up — but don't escalate alone either. Both frameworks are backed by regulatory bodies that exist specifically to take complaints from individuals whose requests were ignored or mishandled. If a company doesn't respond at all, or responds in a way that seems to ignore its obligations, that's generally the point to file a complaint with the relevant regulator or consult a lawyer, rather than trying to force the issue yourself. Does sending this request guarantee the data disappears everywhere? No — and it's worth being honest about that going in. A deletion request can remove data from a company's active systems and account records. It can't retroactively edit things like cached search results, screenshots other people have taken, or — in the case of AI companies — necessarily scrub a model that's already been trained on a dataset that included your information. Deletion requests are a real, meaningful tool, but they're not a surgical "undo" button for everything that exists about you online. What if the company says my data falls under an exception? Ask them to specify which one, in writing. The template above already requests this. A vague "we can't delete this" response isn't really a complete answer under either law — a specific company should be able to point to a specific reason. If the explanation still doesn't add up, that's a reasonable point to involve a lawyer or file a complaint with a data protection authority rather than arguing it out yourself. The AI-specific version of this same request Everything above is the general-purpose version of a deletion request — it works whether you're asking a retailer, a data broker, or a social app to delete your information. But if what you're actually trying to deal with is something an AI assistant says about you, the request looks a little different, because the data involved (training data, not just an account record) behaves differently. We've written a separate piece specifically on how to request removal of personal information from ChatGPT, which walks through that angle in more detail. Either way, sending a request is only half the picture — the other half is checking whether it actually changed anything. You can see whether answers change after a removal by running a free check of what AI assistants currently say about your name, so you're not left guessing whether your request made a difference.
How to Build a Personal Brand That AI Assistants Get Right
For the last decade or so, "personal branding" has mostly meant one thing: controlling what shows up when someone Googles you. Get a clean LinkedIn, maybe a personal website, push down any embarrassing old content, and you were more or less done. That advice still matters. But it's no longer the whole picture, because more people aren't searching for you the old way anymore — they're just asking an AI assistant directly, in a sentence, the way you'd ask a knowledgeable friend. Personal branding used to mean Google. Now it also means ChatGPT. The old version of being "findable" was a list of blue links. Someone typed your name into a search box, scanned ten results, and formed an impression from whatever they clicked. You had some control over that: you could claim your LinkedIn, build a simple website, get a profile on an industry directory, and roughly shape the top of the page. The new version is different in a way most people haven't fully registered yet. Someone types "who is Maria Lopez, the architect in Denver" into ChatGPT or Gemini, and instead of ten links to sort through, they get one confident paragraph. There's no scrolling, no comparing sources, no clicking through to check. Whatever the assistant says is, for that person, simply what's true about you in that moment — even if it's incomplete, outdated, or quietly mixed up with someone else who shares your name. These two versions of findability overlap — a lot of what makes you legible to a search engine also helps an AI model learn about you, since both ultimately draw on text that exists about you somewhere on the web. But they're not identical. A search engine shows its sources and lets a person judge for themselves. An AI assistant compresses everything into a single answer and states it with the same even, confident tone whether it's right or wrong. That's the part most people haven't thought about yet, and it's exactly why it's worth a few minutes of attention now, while it's still a newer problem rather than an obvious one everyone's already solved. If you want to see this gap for yourself, it's worth checking directly rather than guessing. You can measure how well AI knows you with a free check that asks several AI assistants what they know about your name and shows you the answers side by side — a quick way to see whether the version of you that AI gives out actually matches reality. The foundational habits that help with both The good news is that you don't need two separate strategies — one for search engines and one for AI assistants. The habits that make you legible to one largely make you legible to the other, because both are ultimately reading the same web. Pick one name format and stick to it everywhere If you publish as "David T. Chen" on your company bio, "Dave Chen" on LinkedIn, and "D. Chen" on a conference speaker page, you've effectively split one person into three faint signals instead of one strong one. Both search engines and AI models rely heavily on exact-match and near-match text to connect scattered mentions of you into a single, coherent profile. Pick one version of your name — first name, last name, and a middle initial if your name is common enough that mix-ups are likely — and use it consistently across your bio, your social profiles, your byline on anything you write, and any directory listing you control. This matters more the more common your name is. "Maria Lopez" is going to share search real estate and training data with a lot of other Maria Lopezes; "Maria T. Lopez, architect" narrows things considerably and gives both a search engine and an AI model a cleaner thread to follow. Build one or two strong anchor profiles instead of fifteen weak ones It's tempting to think that being everywhere — a profile on every platform, a presence on every network — adds up to a stronger presence. In practice, the opposite is often true. Fifteen thin, half-updated profiles scattered across the internet are weaker signals than two or three that are genuinely complete, current, and consistent. A well-maintained LinkedIn and a simple personal website, both kept up to date, tend to do more work than a dozen abandoned accounts with outdated job titles still listed. Pick the one or two profiles that matter most for how you want to be known professionally, and actually keep them current. When your job changes, update them the same week, not the same year. Say exactly what you do, not vaguely what field you're in Vague bios give both search engines and AI models very little to grab onto. "Works in the food industry" could describe thousands of different jobs and doesn't give anything distinctive to repeat. "Pastry chef in Austin" is specific, concrete, and repeatable — it's the kind of detail that gets picked up accurately and echoed back correctly, because there's only one clear thing it could mean. The same logic applies to any bio you control: your website, your LinkedIn headline, your conference speaker bio, your guest-post author blurb. Specific beats impressive. "Helps small retail businesses set up their first e-commerce store" will serve you better, in both search results and AI answers, than "entrepreneur and business strategist." The habits that build real signal over time Foundational habits keep the record straight. But to actually shape what gets said about you — rather than just making sure what's said is accurate — you need a steady accumulation of real, citable content about you out in the world. This is where a lot of people instinctively reach for social media, and it's worth being direct about why that often doesn't work the way they expect: a lot of social media content either isn't indexed the way a news article or an interview is, or it's indexed but treated as low-signal chatter rather than a durable reference. Posting more on a platform you already use heavily can build an audience, but it doesn't necessarily build the kind of durable, citable record that search engines and AI training data treat as authoritative. What tends to work better is the slower, less glamorous stuff: Genuine media mentions. A local news story, an industry trade publication, a podcast appearance — these get indexed, linked to, and often cited precisely because they're published by outlets with their own established credibility. Interviews and guest contributions. Being quoted as an expert in someone else's article, or writing a guest post for a publication in your field, creates a piece of independent, third-party content that didn't come from you promoting yourself — which tends to carry more weight than self-published content. Speaking engagements. Conference talks, panels, and webinars often generate their own write-ups, recordings, and speaker-bio pages, each one another data point reinforcing the same specific, accurate description of who you are and what you do. None of this needs to happen all at once, and it doesn't require chasing every opportunity that comes along. A handful of genuine, well-placed mentions over a year or two does more for how AI assistants and search engines describe you than a much larger volume of posts that never leave a single platform. It also doesn't mean abandoning social media entirely — it's a fine way to build an audience and stay visible to people who already know you. It just shouldn't be the only thing you're relying on if the goal is shaping how you're described to someone who's never heard of you before and is asking an AI assistant cold. There's also no need to manufacture opportunities that don't fit you. If you're not a natural public speaker, skip the conference circuit and lean harder into guest writing or being a source for journalists in your field instead. The goal isn't volume for its own sake — it's a small, steady accumulation of content that's genuinely about you, written by someone other than you, that a search engine or an AI model has reason to treat as credible. If you want a deeper look at exactly what shows up when someone asks an AI model about you specifically, this piece on what ChatGPT knows about you walks through it in more detail. What this looks like in practice Abstract advice is easier to apply with a concrete example, so consider someone like Priya Nair, a UX researcher who spent a few years building her career without thinking about any of this at all. Before: Priya's LinkedIn listed her old title from two jobs ago. Her personal website, built years earlier, still described her as a "junior researcher," even though she'd since led studies at two different companies. Her bio on a conference page from a talk she gave called her "Priya N." for space reasons. None of this was wrong exactly — it was just scattered, outdated, and inconsistent, which meant anyone piecing together a picture of her, human or AI, was working from mismatched fragments. After a cleanup: she standardized on "Priya Nair" everywhere, updated her LinkedIn and personal site to reflect her current title and a specific description — "UX researcher specializing in healthcare apps" instead of a vague "UX professional" — and let her two or three strongest profiles carry the weight instead of trying to maintain accounts on every platform she'd ever joined. Over the following year, she also did two podcast interviews in her field and contributed a guest post to an industry newsletter, both of which got indexed and linked to from other sites. None of that was a dramatic overhaul. It was a few hours of cleanup followed by patient, ordinary professional activity. But when she checked back a year later by asking an AI assistant what it knew about her, the answer was current, specific, and accurate — a direct result of giving both search engines and AI training data a clean, consistent, well-supported signal to work from instead of a messy one. Check in periodically instead of assuming it's working Here's the part that's easy to skip: after you've tightened up your name format, consolidated your profiles, sharpened your bio, and picked up a few genuine mentions, how do you know any of it actually worked? The honest answer is that you check. Periodically — maybe every few months, or after anything notable happens, like a job change, a big media mention, or a move — ask an AI assistant directly what it knows about you. Not as a vanity exercise, but as a practical one: it tells you whether your efforts are actually landing, or whether there's a specific gap still worth addressing. Sometimes you'll find the AI's answer is accurate and current. Sometimes you'll find it's repeating a job you left two years ago, or confidently blending in a detail that belongs to someone else entirely. Either way, you can't fix what you haven't checked. Treat it the way you'd treat checking your own credit report — not something you do once and forget, but a periodic check-in that tells you whether the picture other people are getting matches the one you intend. How often should you actually check? There's no fixed rule, but a reasonable rhythm is a quick check every few months, plus a check right after anything that changes your professional story — a new job, a notable media mention, a move to a new city. Those are the moments when the AI's picture of you is most likely to be stale, and checking right after gives you the clearest read on how quickly things catch up. What if your name is common and you keep getting mixed up with someone else? This is a genuinely different problem from the ones above, and it deserves its own approach rather than just "be more specific" advice. If you share a name with other people — especially someone more publicly prominent, in a very different field, or with a less flattering public record — an AI assistant can blend details from multiple people into one answer without flagging the uncertainty at all. That's a common enough issue that we've written about it separately: see what to do when AI confuses you with someone else for the specific steps that help separate your identity from a namesake's, rather than just hoping the general branding habits above sort it out on their own. Building a personal brand for the AI-search era isn't really a new discipline bolted onto the old one — it's the same fundamentals of clarity, consistency, and real third-party credibility that have always mattered, applied to a medium that happens to answer in one confident paragraph instead of ten links. Get the fundamentals right, check in periodically, and you're already ahead of most people who haven't thought about this angle at all yet.
What Recruiters See When They Look You Up (Including ChatGPT)
You've got an interview coming up. Before you walk in, there's a good chance someone on the other side of the table has already looked you up — and these days, "looked you up" doesn't just mean a Google search anymore. It increasingly means opening an AI assistant and simply asking, "what do you know about [your name]?" That's a faster habit to pick up than a careful search, and it can hand back a different picture than a search engine would — sometimes a less accurate one, delivered with the same confident tone either way. What a thorough look-up actually covers A recruiter or hiring manager doing their homework before an interview typically isn't doing anything exotic. It usually comes down to a handful of predictable steps: Your LinkedIn. Still the first stop for most people in a professional context — your current title, your work history, who you're connected to. A general name search. A plain Google search of your name, sometimes paired with your employer or city to narrow things down. Public social media. Whatever you haven't locked down — a public Twitter/X account, an Instagram that isn't private, old posts that are still indexed. Increasingly, an AI assistant's summary of you. Rather than reading through search results themselves, more recruiters are just asking ChatGPT, Gemini, or a similar tool directly and taking the answer at face value, especially when they're short on time between back-to-back interviews. None of this is new behavior exactly — people have been Googling candidates for years. What's changed is the last item on that list, and it's worth understanding why it's a meaningfully different kind of risk than the others. Think about Maria Lopez, a marketing manager who's about to interview for a new role. If a hiring manager does a traditional search on her, they'll find her current LinkedIn, a conference bio from two years ago, and maybe an old college newspaper mention — a slightly dated but broadly accurate picture that a careful reader can piece together correctly. If that same hiring manager instead just asks an AI assistant "who is Maria Lopez in marketing," they might get a single paragraph that quietly folds in a detail from a different Maria Lopez entirely — a different city, a different industry, even a different professional reputation — stated with exactly the same confidence as everything that's actually true. Why the AI version is a different kind of risk When a recruiter does a careful Google search themselves, they're looking at a page of results and applying their own judgment — they can see there are three different people named David Chen, notice that one article is from eight years ago, and draw reasonable conclusions about which result is actually relevant. When they ask an AI assistant instead, they get a single paragraph, stated confidently, with none of that visible sorting. A few specific things can go wrong in that paragraph: Blended identities. If you share a name with someone else — especially someone with a more prominent or more public online presence — an AI assistant can merge details from both of you into one answer without flagging any uncertainty. Outdated information. A job you left two years ago, a city you moved away from, a title you've since been promoted out of — AI answers aren't always working from the most current version of you, and there's no visible timestamp to tip off whoever's reading it. Confident mistakes. Sometimes an answer is just wrong, in a way a careful human doing their own search might catch by noticing a mismatched detail, but that reads as perfectly plausible when stated in one smooth, assured paragraph. A recruiter who's inclined to double-check won't take the AI's word as final — but plenty won't double-check, especially when the answer sounds specific and confident. That's what makes this worth a few minutes of attention before a high-stakes interview: it's a real, distinct risk sitting alongside the more familiar ones, not a dramatic one, but one most job seekers haven't thought to check at all. A short checklist before your next interview None of this requires an overhaul. A focused twenty or thirty minutes before a high-stakes interview covers most of it. Clean up anything genuinely unprofessional that's public. Not everything — you don't need to scrub your personality out of your online presence — but anything you'd be uncomfortable having come up mid-interview is worth privacing or removing if it's within your control. Make sure your LinkedIn and any public profile reflect your current role accurately. An outdated title or employer is an easy, avoidable mismatch between what you say in the interview and what shows up when they check. Do a plain Google self-search. See what actually comes up when someone searches your name the ordinary way. If you haven't done this in a while, here's a guide on how to Google yourself properly — it covers more than just typing your name into the box. Take a broader look at your full digital footprint, not just the top search results. A thorough recruiter sometimes digs past the first page — old forum posts, a dormant account, a tagged photo from years back. This guide to checking your digital footprint walks through how to see what's actually out there beyond the obvious first search. Specifically check what an AI assistant says about you. This is the step almost nobody takes yet, which is exactly why it's worth taking. Ask ChatGPT, Gemini, or Claude directly what they know about you, or use a free tool that checks all three at once so you're not repeating the same question three separate times. You can check your name before the interview and see, in one place, what each assistant currently says. What if the AI assistant gets something wrong about you? First, figure out whether it's a mix-up with someone else, an outdated detail, or a plain error — the fix is a little different for each, and it's worth knowing which one you're dealing with before the interview rather than during it. At minimum, knowing in advance means you won't be caught off guard if an interviewer references something that doesn't sound like you — you can calmly correct it instead of looking surprised or defensive. There's no promise that a single check will clean up every stray mention of you online, especially anything baked into how a model was trained — but knowing what's out there ahead of time is still far better than finding out mid-interview that an assistant confidently told your interviewer something untrue. Consider David Chen again, this time applying for a project management role. If he checks ahead of time and finds that an AI assistant is describing a job he left eighteen months ago, he can simply make sure his LinkedIn and any other public profile are current before the interview — a small fix that closes the gap for most future lookups, even if it doesn't retroactively change what a model already learned. If instead the issue is a mix-up with a different David Chen, the fix looks different: adding a middle initial or a location to his professional bios, so there's a clearer, more specific thread for both search engines and AI tools to follow going forward. Is this really worth checking for an entry-level interview, not just an executive search? Arguably more so. The share of recruiters and hiring managers who use AI tools as part of candidate research is widely reported to be rising, and it's not limited to senior searches — a hiring manager screening a stack of applications for an entry-level role is just as likely to ask an AI assistant a quick question about a candidate as someone running an executive search. The checklist above takes the same twenty minutes either way, and the AI-specific step takes less time than the others. You can't control everything a recruiter sees, but you can control this part There's no checklist that guarantees a recruiter forms exactly the impression you want. You can't see every tab they have open, and you can't fully predict what a search engine or an AI assistant will decide to surface on a given day. What you can do is make sure the version of you that's easy to find — on LinkedIn, in a plain search, and now in an AI assistant's answer — is accurate and current. That last part is still a five-minute step most job seekers haven't thought to take, which is exactly what makes it worth doing: it's not crowded advice yet, and a few minutes spent on it before your next interview is a small, cheap way to close a gap most people haven't even noticed is open.
Subscribe to our newsletter
We send helpful materials once a month, along with information about discounts and special offers. No daily spam
We use your email only to send this newsletter. You can unsubscribe at any time from the link in every email. See our Privacy Policy.
How to unsubscribe from the mailing list?