Wikipedia Readers Now Look Up AI Models and Assistants 2.9 Times as Often as Search Engines, a Flip That Happened in the First Full Month After ChatGPT Launched

Summary

Picture the owner of a small accounting practice who notices that new clients now mention using an AI assistant to find her, or to check her before calling. A few years ago they would only have mentioned Google. Wikipedia reading shows where public curiosity has moved.

We grouped English Wikipedia articles into two clusters: one about AI models, chatbots and assistants, and one about search engines. In the past 12 months, readers looked up the AI cluster 2.92 times as often as the search engine cluster.

The switch happened fast. Search engines led by a wide margin through November 2022. In December 2022, the first full month after ChatGPT's launch, AI articles drew 2.77 times as many views, and they have led every month since.

What we measured

We built two lists of article titles and summed their desktop and mobile web views by month from September 2021 to September 2026. New AI articles join the AI cluster from their first month of traffic, which is part of how public interest shows up.

We report cluster totals, ratios and desktop and mobile splits. We do not report interest in any single company or product, and we make no comparison of the products themselves.

AI articles now draw almost three readers for every one on search engines

Annual views and ratio of the two clusters:

Search engine reading was steady through the whole period, between 19.7 million and 22.4 million views a year. The change came almost entirely from new interest in AI.

The flip happened the month after ChatGPT launched

Monthly views around the crossover:

The ratio hit its highest monthly level in April 2023 at 7.13. It has stayed above 1 in every month from December 2022 through September 2026, with its lowest point in August 2026 at 1.45, a month when search engine pages spiked.

The gap is a little wider on desktop

Ratio of AI to search engine views by site:

Desktop, where people work, has led the shift in every year. In 2024 the mobile ratio fell to 1.47 while desktop held at 3.32.

What this means

Public attention has moved toward AI tools, and that matters for how customers find you. Search engines are still used heavily, but more of the public's curiosity now goes to AI, and likely more of the first questions too. A business that only tracks its Google rankings is missing part of the picture. Good AI search optimization services track the assistants as closely as the rankings.

Check how AI assistants answer the questions your customers ask, and whether they mention you. Then fix the gaps with clear, accurate information on your site and listings. A free AI visibility audit shows where you stand in ChatGPT and the other assistants before you spend time on ChatGPT SEO fixes.

Limits

Wikimedia pageview counts include some automated traffic that its filters do not catch, and Wikimedia changed how it detects bots in 2025, which lowers counted views from that point on. Some of the change in any period may reflect better bot filtering rather than fewer human readers.

We count views of titles that had at least 5 views in a given month on a given site, which drops a very small tail of rarely read pages.

Wikipedia reading measures curiosity about a topic, not how many people use a product.

The AI cluster grows as new articles are created (from 14 articles in early 2022 to 41 by 2025), which is part of the trend but also inflates later totals.

Search engine pages include company pages, which also draw readers for business news.

Frequently asked questions

Related studies

• Wikipedia readership in the AI answer eraSEO vs GEO on Wikipedia: SEO Page Views Down 51.9%
• Wikipedia readership in the AI answer eraHome System Pages on Wikipedia Lost 50.4% of Readers
• Wikipedia readership in the AI answer eraLegal Concept Pages Lost 46.9% of Wikipedia Readers
• Wikipedia readership in the AI answer eraFinance Explainer Pages on Wikipedia: Median Drop of 49.9%