For years, brands have made media placement decisions based on a publication’s domain authority, readership numbers, and name recognition. None of those metrics measures how often a publication actually appears when someone asks an AI engine for a recommendation.

That gap is what the Baden Bower AI Visibility Index 2026 was built to close. The study ran 20 buyer-intent questions through six AI engines, ten times each, and logged every source those engines cited in their responses.

The six platforms covered were ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews, and Microsoft Copilot. Across 1,200 total observations, the methodology captured 12,040 individual citations. Every citation was scored and attributed to its source publication, producing a ranked index of which outlets AI engines actually quote when buyers are making decisions.

The methodology was designed to reflect the full AI search landscape rather than a single platform. Because citation behavior varies across engines, each publication was scored on its average performance across all six simultaneously, producing rankings stable enough to compare across the category.

Who Wins in AI Answers

Forbes ranked first, scoring 92 out of 100 on the index’s AIO Citation Score. The score is an index number, not a percentage: each publication’s raw citation share across all six engines is normalized so that the top-ranked source scores 92, making it easier to compare the distance between any two publications on a single scale. Forbes’s underlying citation share was 13.8 percent of all citations logged, roughly a third higher than the 10.5 percent recorded by the second-ranked source, Business Insider.

That gap between first and second place is notable in itself. TechCrunch followed at 53.5, Bloomberg at 48.9, and Wikipedia at 48.4. Entrepreneur scored 40.7, Inc. at 35.4, Fast Company at 34.3, and Reddit at 33.2.

The result that will surprise most media strategists sits at rank ten. Yahoo Finance, one of the most widely distributed financial news platforms in the world, scored just 28.8, placing it below seven editorial outlets, Wikipedia, and the community platform Reddit, which scored 33.2. Distribution scale, in other words, does not translate directly into AI citation frequency. The engines appear to weigh content type and editorial origin more heavily than reach alone.

The top ten is dominated by editorial sources, publications where stories are written, edited, and placed through journalistic relationships rather than distributed via press release wires. Seven of the top ten-ranked sources are editorial outlets. The three exceptions are Wikipedia and Reddit, which carry a mix of content types, and Yahoo Finance, which functions primarily as a distribution platform.

“Distribution gets a brand in front of readers. It does not get a brand into AI answers,” said AJ Ignacio, CEO of Baden Bower. “The engines are making a different judgment about what counts as credible. This index shows which sources pass that test and which ones don’t, regardless of how widely they circulate.”

Three Patterns That Define AI Citation Behavior

The index surfaces three structural patterns that hold consistently across all six engines and across all query types tested.

The first is the editorial advantage. Authored editorial coverage, stories placed through editorial relationships rather than distributed as press releases, was cited 2.3 times more often than wire-distributed coverage of the same brands.

Wire services, including AP News, Reuters, and PR Newswire, all appeared in AI answers, but at significantly lower rates. PR Newswire scored 12.5, the third lowest of any source in the study. The implication is that the channel through which coverage is produced and distributed affects AI visibility independently of which publication carries it.

The second pattern concerns publication tiers. Flagship editions carry the heaviest citation weight: Forbes flagship scored 92, while its country editions scored 17. Both appeared in results from all six engines, and the country editions still outscored press release wires such as PR Newswire. The data points to a tiered approach rather than a single play: flagship placements anchor a brand’s AI visibility, edition-level editorial coverage broadens it, and wire distribution alone delivers the least.

The third pattern is recency. Coverage published within the previous 12 months was cited 2.9 times more often than older coverage of the same companies. The effect held across all content types and all engines tested. Historical coverage from the same publication, even coverage that performed well at the time of publication, contributes materially less to current AI visibility than recent placements do.

The Unexpected Performers

Two sources in the index outperformed expectations by a significant margin. Wikipedia ranked fifth overall at 48.4, above Entrepreneur, Inc., Fast Company, and every wire service in the study. Reddit ranked ninth at 33.2, ahead of Yahoo Finance, AP News, Vogue, and Vanity Fair.

Wikipedia’s performance reflects its established role as a primary reference source in AI training data. Its content is structured, comprehensive, and frequently cited across all major AI platforms. For brands that have invested in editorial media coverage but neglected their Wikipedia presence, the index suggests a meaningful gap in AI visibility strategy.

Reddit’s ranking is a more recent development. The platform’s community-generated content, particularly in threads structured around questions and recommendations, closely mirrors the format of buyer-intent queries. As AI engines increasingly retrieve content based on question-answer structures rather than purely on domain authority, forum-based platforms appear to be gaining citation weight.

Vogue ranked 12th at 25.1, and Vanity Fair ranked 13th at 24.6, placing both behind Reddit and Yahoo Finance in AI citation frequency. Both publications carry high editorial prestige and significant cultural authority, but their citation rates in buyer-intent queries reflect the more specialized nature of their content relative to business and technology publications.

The Emerging Practice of AI Visibility Planning

The index arrives at a moment when a growing number of brands are beginning to plan media activity specifically around AI discoverability, a practice the industry has started calling Generative Engine Optimization, or GEO. The premise is that appearing in AI-generated answers to buyer questions is becoming as commercially important as appearing in traditional search results, and that the two require different, though overlapping, strategies.

The data from this index gives brands a concrete basis for evaluating media placements against AI visibility outcomes before committing to them. A placement in a flagship publication, secured through an editorial relationship and published in the past 12 months, sits at the top of the citation hierarchy. A press release distributed via wire service, with no editorial placement, sits near the bottom.

Baden Bower has built a service practice around this category, tracking client citation rates across AI platforms as a core reporting metric. The company’s 2026 research program, which includes four independent studies surveying more than 4,800 respondents, represents one of the more extensive commissioned datasets on earned media’s commercial impact currently available in the PR industry.

The full AI Visibility Index 2026, including the complete publication rankings, methodology, and engine-by-engine breakdowns, is available here.