Marketers have a new force to reckon with, and it lives inside the AI assistants customers now use to research brands. New analysis from Seer Interactive, built on 2.7 million data points gathered during the 2026 Winter Olympics, suggests that AI search tools don’t simply report the current facts about a company. They retell a story that was set in place long before the user ever typed a question.
The research team, led by Seer’s VP of Analytics John Lovett, used the Games as a controlled testing ground: predictable storylines, constant breaking news, global coverage, and clear, verifiable outcomes. Over a nine-week period they asked the same questions every day across six major platforms, including ChatGPT, Gemini, Google AI Mode, AI Overviews, Perplexity, and Meta AI, tracking how each system surfaced, cited, and suppressed information.
What the researchers call “narrative gravity”
A consistent pattern emerged. Ahead of an event, a consensus narrative would form around a favored athlete or team. When the real result broke from that expectation, the AI systems often kept echoing the pre-event story anyway, in some cases describing a favorite as the winner when they had actually lost. The established narrative, in other words, exerted a kind of gravity that the latest facts struggled to overcome.
For brands, the implications are direct. If the dominant story about your company was shaped months or years ago by a Glassdoor review, a news cycle, or an analyst report, an AI assistant is likely to keep telling that version even after the reality has moved on. A customer who asks ChatGPT or Perplexity about your business doesn’t just receive today’s facts; they inherit an accumulated storyline.
What to do about it
The practical takeaway, presented at Seer’s AI for B2B Marketers Summit, is that reputation management now has an AI dimension that operates on a longer timeline than traditional PR. Outdated narratives need to be actively countered with fresh, consistent, well-sourced signals across the channels these models draw on, rather than a single announcement. Brands that wait for AI systems to “catch up” on their own may find the old story keeps surfacing long after it stopped being true.
