Try this one evening. Type your full name and city into a search engine and scroll past the first page. For most people, it doesn’t take long to find an old address, a phone number, maybe a list of “possible relatives” with ages beside them. Nobody asked your permission to put it there, and you probably never noticed it arrive.
Privacy used to be about walls. You closed a door, locked a drawer, lowered your voice. Now your information drifts between apps, ad networks, smart speakers and data brokers, and lately AI systems can stitch it all together faster than any private investigator.
Whether your data is out there isn’t really in question. It is. What matters is how much say you still have once it starts moving.
The first step is working out how it got public in the first place. Services like ClearNym.com track people-search sites and broker databases and file removal requests, both for individuals and for companies trying to protect their staff. Do it yourself or pay someone, but expect it to be ongoing. Delete a listing from one site and it can quietly pop up on another a month later.
AI didn’t invent the problem. It scaled it.
A single fact about you is mostly harmless. Your ZIP code. The gym you go to. The time you usually buy coffee. Put twenty of those together, though, and someone can guess your routine, who you live with, even what you might be worried about health-wise.
That kind of stitching used to take real effort. Now it’s cheap.
The US National Institute of Standards and Technology has been fairly blunt about it. NIST says AI creates new re-identification risks, because it can match “anonymous” records back to real people across different datasets. It can also amplify behavioural tracking and surveillance. And training a model can leak data, meaning personal details fed in can sometimes come back out.
To be fair, no single system knows everything about you. Your running app has your routes. Netflix knows what you binge. Some ad network has your phone’s identifier. The worry is what happens when something connects them.
“If you’ve nothing to hide…”
You’ve heard that line. It’s lazy.
Privacy is written into Article 12 of the Universal Declaration of Human Rights, which protects people from arbitrary interference with their home, family and correspondence. UN bodies have said more than once that the same right holds online. It’s about dignity and being able to think things through without an audience. It has nothing to do with guilt.
Governments complicate this. Police and intelligence agencies sometimes have legitimate reasons to get data, under court orders and rules. But the fights over the NSA’s bulk phone-records programme taught an uncomfortable lesson: metadata alone (who you called, when, for how long) can reveal an awful lot, even if nobody ever listens to a single call. A warrant for one case doesn’t cancel everyone else’s expectation of privacy.
How a few clicks turn into a profile
Here’s what people miss about AI and data. It doesn’t just store what you told it. It guesses.
Buy a pregnancy test, a certain vitamin and some unscented lotion, and a retailer’s model may decide you’re expecting before you’ve told your mum. Location pings around a clinic tell their own story. Facial recognition can put a name to a face in a crowd photo.
Cambridge Analytica is still the textbook case. The FTC found the firm used deceptive practices to collect data from tens of millions of Facebook users and then profiled and targeted them politically. Most of those people never knowingly agreed to any of it. They’d just taken a quiz, or had a friend who did.
Where it leaks | What’s at stake | Quick fix |
| Phone apps | Constant tracking | Strip unused permissions |
| Smart speakers and TVs | What happens at home | Turn off data sharing you don’t need |
| People-search sites | Address, phone, relatives | Request removal |
| AI chatbots | Whatever you paste in | Think before you upload |
| Ad networks | Your interest profile | Turn down personalisation |
Opting out is boring. Do it anyway.
The manual way goes like this. Find the site’s opt-out page, which is often buried in the footer. Search for your record. Confirm your email. Click a verification link. Submit the removal. Then repeat for the next site, and the next.
It’s tedious. It still matters. Removing listings makes it harder for a stranger, or an ex, to turn up at your door. It shrinks the pool of data feeding ad and AI profiles. And it gives you a record to point to when the listing comes back, because it often does. Brokers refresh their databases constantly from other sources, which is why monitoring services exist at all. ClearNym, for one, scans broker and people-search sites, files the requests and keeps watch for re-listings. No honest provider will promise a permanent wipe, though, because the data keeps coming from somewhere.
There’s genuinely good news if you live in California. In January 2026 the state launched DROP, a free tool that sends one deletion request to every registered data broker at once. Brokers have had to start acting on those requests since August 1, 2026. Public records aren’t covered, and it only helps Californians. Still, it’s the first time a single form has reached the whole broker industry, and other states are watching.
Why your employer should care too
Think about what an attacker can do with a CFO’s home address, an IT admin’s mobile number and the names of their kids.
That’s the raw material for convincing scams. A text that seems to come from the boss. A call that mentions your daughter’s school. Researching a target used to take an attacker hours. With broker data and a chatbot, it takes minutes. More companies now pay to scrub employees’ and executives’ details from broker sites for exactly that reason. It sits alongside phishing training and access controls, not instead of them.
What’s actually in your hands
Quite a lot, really. Keep your home address and phone number off public profiles. Skim the privacy policy before uploading anything personal. It takes two minutes and you’ll be surprised how often it changes your mind. Go through your phone’s app permissions and switch off location for anything that doesn’t obviously need it. Turn down ad personalisation in your Google, Meta and phone settings. Request removal from people-search sites. And recheck all of it after big software updates, since settings have a habit of resetting themselves.
None of that makes you invisible. It does make you a lot harder to profile.
Be especially careful with AI chatbots. Back in 2023, Samsung clamped down on staff using generative AI after engineers pasted confidential source code into one. Ordinary users are no better. Look at publicly shared chats and you’ll find people who share sensitive personal details, such as phone numbers, emails and health worries, without a second thought. At work it’s the same story, with staff who paste company secrets into chatbots because it’s quick. My rule of thumb: if I wouldn’t email it to a stranger, I don’t paste it into a tool whose data handling I haven’t checked.
The law still has teeth (sometimes)
Regulation matters more as AI spreads, not less.
In Europe, the GDPR says companies need a lawful reason to process your personal data. If that reason is consent, it must be freely given, specific, informed and unambiguous, and withdrawing it has to be as easy as giving it. In practice, that’s why so many EU cookie banners now offer a real “reject all” button.
The trouble is that one app on your phone might quietly pass data to a dozen vendors you’ll never meet: cloud hosts, analytics firms, ad exchanges. Clicking “accept” on the first screen tells you almost nothing about where things end up. Accountability has to follow the data down that chain, and regulators are only slowly catching up.
Some of this is out of your hands
Honestly, plenty of it is. Property records and court filings are public whether you’re on social media or not. Cameras watch most high streets. Brokers buy data from sources you’ve never heard of.
That’s where laws and corporate policy have to step in, because no individual can audit every database. The money matters as well. Your data is worth something to someone, and as long as that’s true, companies will want more of it. Only real rights, actually enforced, push back on that.
There’s a subtler shift too. Privacy researchers talk about practical obscurity: a fact that’s technically public but hard to dig up, like a 20-year-old court record sitting in a county basement. AI search quietly erases that. Things that were “public” only in theory are becoming public in practice.
So, is privacy dead?
No. But it’s no longer something you set up once and forget. It’s more like keeping a garden.
Check what’s already out there. Share a bit less. Trim app access. Lock down your accounts with strong passwords and two-factor login. Request broker removals and keep an eye out for listings that reappear. Delete old accounts you don’t use. Then come back in a few months and do it again.
AI has made the risks bigger and quicker. It hasn’t taken away your choices. What you disclose, what you allow and what you ask to have removed still add up to a meaningful amount of control, far more than shrugging and assuming it’s all gone anyway.
Quick answers
Q. Can AI figure out things I never told it?
Yes. It infers from patterns, and it’s sometimes wrong, but often close enough to matter.
Q. If I delete an app, is my data gone?
Usually not. Deleting the app and deleting your account are two separate things. Do both.
Q. Does incognito mode make me anonymous?
No. It just stops your own browser saving history. Websites and your network can still see you.
Q. Are public records really a privacy problem?
They can be, once broker sites scoop them up and make them searchable in seconds.
Q. Can I be completely private online?
Realistically, no. You can still be a lot less exposed than you are today.
Related: Is ChatGPT Safe in 2026? Privacy, Security & What Changed
