Why Does AI Repeat an Old Complaint about My Company?

AI keeps bringing up a problem you resolved long ago, or calls you untrustworthy? Here are direct answers on why AI repeats old complaints, why it sometimes invents negatives, how AI reputation differs from SEO and PR, and how to get AI to represent your standing fairly.

If ChatGPT or Perplexity keeps bringing up a problem you resolved long ago, or frames your company as untrustworthy, you are seeing a Reputation problem. Here are the questions brands ask most about AI and trust, with direct answers. For the full picture, see the Reputation pillar guide.

Why does AI repeat an old complaint about my company?

Because the complaint is still well represented in the sources the engine reads, and the resolution is not. Engines synthesize from what exists online, so a loud old controversy, a cluster of dated negative reviews, or a single viral incident can dominate long after reality has changed. If there is little recent, credible signal to outweigh it, the engine keeps surfacing the old story. The fix is to make the current, accurate picture the one engines have the most of.

Why does AI say my company is unreliable when our reviews are good?

Often because the engine is weighting a few visible negatives, or older sources, more heavily than your recent positive reviews, or because your strong reviews live on platforms it is not drawing from. It can also stem from mistaken identity, where another company's problems attach to yours. Maintaining recent reviews on the platforms engines trust, and keeping your company facts consistent, helps the accurate picture win.

Why does AI think my company shut down?

Usually because your current signals are sparse and something looks like an ending. A discontinued product mistaken for the whole company, an abandoned profile, a similarly named business that did close, or a quiet news period can all lead a model to guess that you are gone. Keeping your company status and facts current and consistent everywhere gives the engine a clear answer instead of a damaging guess. We cover this worst case in When AI Says You've Shut Down.

Is AI making up negative things about my brand?

Sometimes, yes. When a model lacks clear, current trust signals, it can state a confident falsehood, an invented breach, a lawsuit that never happened, the same gap-filling that causes errors in other categories. It is most likely where your reputation footprint is thin. Publishing clear, current trust signals is the most reliable way to remove the ambiguity it fills with guesses.

Can I get AI to remove a negative claim?

You cannot edit an engine's answer directly, but you can change what it reads. Engines update as their sources do, so resolving and responding to the underlying complaints, earning current coverage, and publishing your own factual account shifts the weight of evidence over time. Trying to bury legitimate criticism rarely works and can backfire. Correcting the record and outweighing the old narrative does.

How is fixing this different from SEO or PR?

It overlaps with both but is its own discipline. Traditional SEO ranks pages, and PR shapes coverage, while getting AI to represent your reputation fairly is answer engine optimization (AEO), sometimes called generative engine optimization (GEO): making sure the trust signals an engine synthesizes are current and accurate. You can have good PR and still have AI repeat an old story, because the engine draws on more than your latest headlines.

Does it matter if AI gets one trust detail wrong?

Yes, more than almost anywhere else. A trust answer lands at the moment of commitment and feeds the loyalty loop, so a false or hostile one ends deals and shapes the next customer. In a HiBot audit, a false or materially negative claim caps the answer at 2 regardless of any positives and pulls down your Accuracy Rate. A confident wrong answer about trust is treated as a serious loss, because it is one.

How do I check what AI says about my reputation?

Run trust prompts, "is this company reliable," "what do reviews say," "are they still in business," logged out, across more than one engine, and record every false claim and stale negative. You can structure the check with The Trust-Signal Checklist for AI Visibility. Remember that answers vary from ask to ask, so a single check is a hint rather than a verdict.

If you want to know exactly what AI says about your standing first, a HiBot audit measures your Reputation and Accuracy Rate across engines and ranks what to fix. Download the ANSWER whitepaper to see the method, or request an AI visibility audit at hibot.com.

David Tang
David Tang · Corporate Strategy, New York
David Tang is the CEO and Founder of HiBot and Flevy. Flevy is the world's largest marketplace for business frameworks and templates. Prior to these companies, David worked as a management consultant for 8 years, where he served clients in North America, EMEA, and APAC. He graduated from Cornell with a BS in Electrical Engineering and MEng in Management. LinkedIn →
Connect our AI visibility playbook to your AI