The new gatekeeper

Ask ChatGPT a question about your business and it will answer with total confidence, whether or not it has the faintest idea what it’s talking about.

A woman looking for a florist no longer opens 10 browser tabs and squints her way to a verdict. She asks once, and the chat answers back, certain as a weather forecast. It might be right. It might be repeating something three years out of date. Either way, someone is about to act on it.

That shift is now measurable. According to CO by the U.S. Chamber of Commerce, 84% of people now make buying decisions based on the first suggestion an AI tool gives them. Debra Vilchis, president of Fishman PR, told the outlet that consumers now routinely cut off their Google searches after seeing the AI summary. The search does still happen. The click often doesn’t. The decision comes first.

The discipline emerging around this shift has a name: Generative Engine Optimization, or GEO. A comparative study of AI search and traditional Google search, published on arXiv by researcher Kaiwen Chen and co-authors, found that AI search systems show a systematic and overwhelming bias toward earned, third-party media over content a brand publishes about itself, a sharp contrast to Google’s more balanced mix of sources.

Where SEO was built to earn a spot on Google’s results page, GEO is built to win a mention inside an AI-generated answer, whether or not the person asking ever clicks through to a website at all. Google’s own data on the underlying shift shows how quickly the change has happened; according to Semrush’s analysis of more than 10 million keywords, AI Overviews appeared in just 6.49% of Google search queries in January 2025, climbed to nearly 25% by July and settled at 15.69% by November. Less than a year in, roughly one in six searches was already answered before it was ever clicked.

The scale of what’s at stake is starting to show up in case studies agencies are willing to put their name on. Optimist, a content marketing agency working with B2B technology and software companies, reports on its own site that one client tracked a 4,900% increase in AI-referred revenue, a 49-times jump, over a 14-month engagement, measured through direct attribution from AI referral sessions into closed revenue in the client’s CRM.

Public relations professionals are watching their own role evolve with it. Gini Dietrich, founder and CEO of Spin Sucks and creator of the PESO Model, a communications framework widely taught in PR programs, argues that communicators need to move past vanity metrics and start tracking how often their brand or their experts turn up inside AI-generated summaries.

She’s also blunt about the limits of the technology doing the summarizing, writing that AI is only as good as the data it’s trained on, and that the training data itself is riddled with bias and gaps. The AI isn’t infallible, but it is already making judgment calls about who to mention, caveat or no caveat.

Julia Ruttner, co-founder of Salt + Ruttner Communications, made a related argument in a piece for Forbes, describing a progression from search-engine-driven discoverability, to social-algorithm-driven reach, to a new layer of AI-generated synthesis that now shapes how people find and trust information in the first place.

The pattern she’s describing has a price tag attached, or at least a projected one. Gartner, the research and advisory firm, has predicted that mass adoption of public large language models as a replacement for traditional search will drive a two-fold increase in PR and earned media budgets by 2027.

Not every expert agrees on how long today’s visibility metrics will matter, either. Eli Schwartz, a longtime SEO strategist, predicted in a roundup of AI-search commentary that simple visibility scores, how often a brand shows up in an AI answer at all, will matter a lot less by the end of 2026, as more sophisticated tools emerge to measure whether that visibility converts into anything. The metric is new enough that even the people building it expect it to change shape within the year.

Local chambers of commerce got the memo before most small business owners did. The Lebanon Wilson County Chamber in Tennessee built a seminar around the fact that customers are no longer relying solely on Google, walking members through why some businesses are easier for AI tools to find than others.

Florida’s Apopka Area Chamber published member guidanceexplaining that AI tools now act as digital referral sources, synthesizing information from across the internet rather than simply listing links, and that businesses with thin digital footprints risk being overlooked entirely. Chambers typically build programs about what members are already asking, which means this concern reached Main Street faster than the marketing trade press gave it credit for.

What separates a business the AI mentions from one it skips isn’t size or budget. It’s specificity. Softologics, a small business marketing firm, describes working with a bakery owner who wanted to rank for the word “marketing,” a category the firm compares to a continent rather than a lane. Once the owner narrowed the business to custom cakes and corporate catering in one city, inquiries sharpened, because the site started sounding like a specific, real business instead of a template brochure. AI systems, it turns out, reward exactly the kind of plain-spoken specificity that also makes for better human copy.

The other side of this shift is happening inside PR departments themselves, and not everyone is thrilled about it. As firms lean on AI to draft pitches and monitor coverage, reporters are drowning in the output.

Will Swope, associate director of issues management and monitoring at the National Cattlemen’s Beef Association, predicted that communications teams would need to add more time and resources to AI monitoring just to keep pace. Separately, PRLab’s 2026 industry report found that 59% of PR professionals now rank AI as their top priority, even as journalists report fatigue from a flood of generic, obviously automated pitches. The tools got faster. The bar for sounding like a person who did the research went up, not down.

There’s a version of this story with a big-budget cast too. CeraVe’s “New Face of Legs” campaign, built around a 2021 viral post about NBA star Kevin Durant’s dry skin, generated a 43% sales lift along with more than 4.1 billion PR impressions and 83 million organic video views, CeraVe US general manager Esther Garcia told Marketing Brew in an exclusive.

The three-week campaign leaned almost entirely on social content rather than traditional media buys. That’s not a small business story, but the underlying lesson travels down market: the brands winning attention right now are the ones that can prove impact in terms a machine can parse and repeat, not just terms a human finds persuasive.

None of this replaces the fundamentals. A business still has to be good, and it still has to say so honestly. What’s changed is the audience for that message now includes a layer of AI systems deciding, on a customer’s behalf, whether the business is worth mentioning at all.

Finding out where you stand takes about five minutes: open an AI tool, ask it what it knows about your business or your industry and read the answer the way a stranger would. If it’s wrong, outdated, or silent, it’s not a footnote. That’s the story, and it’s one you can start rewriting today.