
Why AI Hallucinates About Your Company
ChatGPT says your CEO went to Stanford. She went to Penn. Why AI engines hallucinate about brands, and how to close the citation gap before it compounds.

ChatGPT says your CEO went to Stanford. She went to Penn. Why AI engines hallucinate about brands, and how to close the citation gap before it compounds.

AI engines treat online casino queries differently than sportsbook queries. Tighter citation. Harder hedging. More legality framing. The operators that understand the asymmetry will own discovery in a category the licensed sportsbook story does not cover.

ESPN owns the news. Wikipedia owns the record. Reddit owns the takes. The pilot framework for the EPR Sports Citation Index — measuring sports citation share inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.

Perplexity does something the other major AI answer engines do not yet do as aggressively: it surfaces source citations directly inside the answer interface, with hyperlinks, rankings, and a hover-to-preview that reads more like a journalist's sidebar than a search result. For communications teams, this matters in a way that is easy to underestimate.

How AI engines persist healthcare crisis narratives for 12-18 months, replacing the traditional news-cycle half-life. The new response window is 2-4 weeks. Source dominance beats message dominance. Theranos, Purdue, J&J, and prestige-name case examples.

A directional modeling study of how five AI engines — ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews — surface and rank the 15 industries most associated with reputation crisis. Banking #1, Pharma #2, Aerospace #3, Automotive #4, Tech Platforms #5. The sector-side companion to the Corporate Crisis Citation Share Index.

For three decades the healthcare funnel was a long sequence. AI engines have compressed it into one moment. Budget, measurement, organizational structure, crisis posture, and sales enablement all rebuild around the new top of funnel.

Twenty-five prompts. Five engines. One hundred twenty-five outputs. Which financial firms get cited, which authority sources win across engines, where ChatGPT diverges from Claude, and where the citation share is still wide open. The data anchor for the Financial Services pillar.

A buyer asks ChatGPT, "Best [category] vendors," and your brand isn't in the answer. Or worse, it's named once, in a footnote, after three competitors you've outsold for a decade. There are typically four reasons. Most brands assume the wrong one.

The article discusses how AI retrieval systems determine reputation, outlining the key "anchor surfaces" they use to synthesize information. It emphasizes the importance of adapting PR and marketing strategies to these new indexing methods, providing insights into what gets a subject favorably indexed and what doesn't. The text also highlights real-world examples of reputation shifts experienced by public figures and companies due to AI indexing.