You open a new tab. You type your own company name into ChatGPT. And you read a description of a business that ceased to exist three years ago.
Wrong job title. Old tagline. A category you repositioned out of a long time ago. You didn’t write any of that. You never approved it. And here’s the part that should actually bother you — you can’t just log in and fix it.Here’s what most business owners still don’t understand about how AI describes them. Your website isn’t the source of truth. It’s one input among hundreds.
Hi, I’m Jeff Payne. You’re listening to The Jeff Payne Show, Episode #60: Someone Else Owns Your Reputation Now.
AI systems don’t take your word for who you are. They cross-reference everything they can find — old interviews, directories, author bios, press mentions, forum posts — and they build a composite. Think of your website as your opening statement. Everything else mentioning you is a witness — and AI weighs the witnesses more heavily than the statement.
That’s the mechanism behind the Two Currencies idea we’ve talked about before — citations carry more weight than mentions. What we haven’t talked about yet is who actually controls those citations. And the answer is: not you.
Here’s where it gets uncomfortable. AI systems have a rough way of deciding what’s true — repetition. If the same description appears on enough third-party sites, it starts to look authoritative, even if it’s years out of date.
So a business can reposition itself completely, and if 40 old articles still describe the old version, that old version is what keeps surfacing. Not because anyone’s lying. Because nobody updated the record — and updating it was never really the business’s job to begin with.
That’s the part that should reframe how you think about this. It’s not a content problem. It’s a control problem. You’re being described by people who have no obligation to keep your description current.
This isn’t a fringe concern — it shows up in the numbers. Recent citation data across industries breaks down how much of a brand’s AI presence comes from what the brand controls versus what everyone else says about it. In some categories, brand-owned sources account for less than half of the citation mix. In several, they’re barely a third. The rest — earned coverage, social mentions, third-party pages — is entirely outside the business’s control.
Put plainly: for a meaningful share of companies right now, the majority of what AI says about them is being written by people who aren’t them. And once enough businesses figure that out — once they realize their AI visibility depends on pages they don’t own and can’t edit — a predictable thing happens. The people who do control those pages start to notice they’re sitting on leverage. What was once a free correction, a quick editorial update, starts to look like something worth charging for.
Which raises a real ethical line worth drawing here. There’s a difference between correcting something that’s actually wrong — an outdated title, an old company description presented as current — and erasing something that’s simply inconvenient. A discontinued product that once got real press. A quote from an executive who’s since moved on. An honest but unflattering review. None of that stops being true just because it’s no longer convenient to have said.
One should be fixable. The other shouldn’t be for sale.
So where does that leave you? Not with a task list — with a question worth sitting with: If someone searched an AI system for your business today, whose version of you would it find? The one you’d write, or the one that’s been sitting untouched somewhere for years?
You don’t control most of that record. But you can find out what it actually says about you before it becomes a problem you’re negotiating from a weaker position.
The internet has a long memory, and most of it isn’t yours to edit. The businesses that end up in the strongest position aren’t the ones scrambling to correct the record after AI gets it wrong. They’re the ones who already know what the record says.
AI doesn’t just repeat what your website says about you — it repeats what everyone else says, too. As businesses realize how much that third-party record shapes their AI visibility, the people who control it are starting to recognize the leverage it provides.
THE DESCRIPTION YOU NEVER APPROVED
Type a company name into an AI system, and it’s common to get back a description that’s years out of date — an old tagline, a repositioned category, a title nobody holds anymore. Nobody at the company wrote that description. Nobody approved it. And there’s no login that lets them fix it.
That’s because a business’s website isn’t the source of truth AI relies on. It’s one input among hundreds — an opening statement, not a verdict.
REPUTATION LOOKS LIKE TRUTH
AI systems don’t take a company’s word for who it is. They cross-reference everything they can find — old interviews, directories, author bios, press mentions — and build a composite from it. When the same description repeats across enough third-party sources, it starts to read as authoritative, even years after it stopped being accurate.
A business can reposition itself entirely, and if 40 old articles still describe the earlier version, that’s the version that keeps surfacing — not because anyone lied, but because updating the record was never really the business’s job to begin with.
That reframes the problem. It isn’t a content gap. It’s a control gap — being described by people under no obligation to keep that description current.
Your website is your opening statement. Everything else that mentions you is a witness — and AI weighs witnesses more heavily.
This is one of the best written, most informative email newsletters I’ve ever read!
Thank you for sharing your expertise, and hope you’re doing well.
Beth Zimmerman
Founder, Executive Director
Who ACTUALLY controls the record?
Recent citation-mix data across industries shows how much of a brand’s AI presence comes from sources the brand controls versus everything else. In some categories, brand-owned sources make up less than half the citation mix; in several, barely a third. The remainder — earned coverage, social mentions, third-party pages — sits entirely outside the business’s control.
For a meaningful share of companies right now, much of what AI says about them is written by people who aren’t them.
For a meaningful share of companies right now, much of what AI says about them is written by people who aren’t them.
When correction becomes leverage
As more businesses realize how much their AI visibility depends on pages they don’t own, a predictable dynamic emerges: the people who control those pages start to recognize they’re sitting on leverage. Nick Leroy of SEOForLunch has written about this as an emerging risk worth watching — publishers who once offered free corrections are starting to treat them as billable.
That raises a real distinction worth holding onto: correcting something that’s actually wrong — an outdated title, an old description presented as current — is different from erasing something that’s simply inconvenient. A discontinued product that once got real press. A quote from an executive who’s since moved on. An honest but unflattering review. None of that stops being true just because it’s no longer convenient to have said.
One should be fixable.
The other shouldn’t be for sale.
What is actually worth checking?
The useful question isn’t a task list — it’s a single one worth sitting with: if someone searched an AI system for a business today, whose version would they find? The one the business would write, or the one that’s been sitting untouched somewhere for years?
Most of that record isn’t something a business controls. But it is something a business can find out — before it becomes a problem being negotiated from a weaker position.
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
Subscribe and Share – WE APPRECIATE YOUR SUPPORT
