AI Overviews Cut Wikipedia Search Traffic by Five Percent

How Researchers Measured the Referral Decline

A study from the University of Washington found that Google's AI Overviews feature reduced search referrals to Wikipedia by approximately five percent. The research analyzed web traffic patterns before and after the rollout of AI-generated summaries in search results. The decline was measured across multiple language versions of the encyclopedia site. Researchers noted the drop occurred despite stable overall search volumes for related queries. Google has questioned the methodology used to track referral changes. The findings suggest AI Overviews may be satisfying user queries directly on the search page. This reduces the need for users to click through to external sources like Wikipedia. The study highlights growing concerns about how AI features affect traffic to knowledge-based websites.

The University of Washington team compared click-through rates from Google search to Wikipedia pages during periods with and without AI Overviews active. They isolated traffic from informational queries where AI summaries are most likely to appear. The analysis controlled for seasonal trends and major news events that could skew results. Researchers used anonymized search log data from a large sample of users. They focused on pages where AI Overviews commonly summarize content, such as historical events or scientific concepts. The five percent figure represents an average across observed categories. Some topics showed larger drops, while others saw minimal change. The team emphasized their metric tracks only search-driven visits, not direct or social media traffic.

Why Google Challenges the Findings

Google has disputed the referral metric used in the study, arguing it does not fully capture how users interact with AI Overviews. The company claims internal data shows no significant impact on traffic to authoritative sites like Wikipedia. Google maintains that AI Overviews are designed to complement, not replace, visits to source pages. They point to links within the overviews that direct users to original content. The tech giant says user behavior studies indicate people often click through for deeper information after viewing summaries. Google also notes that Wikipedia itself uses AI tools internally, which may complicate external assessments. The disagreement reflects broader tensions between publishers and platforms over how AI affects content visibility.

What Does This Mean for Online Knowledge Sharing?

The referral decline raises questions about the long-term sustainability of ad-supported and donation-funded knowledge platforms. If AI Overviews continue to grow, sites relying on search traffic may need to adapt their visibility strategies. Wikipedia, which depends on public contributions and donations, could face reduced exposure to new readers. However, the platform’s strong brand recognition may help mitigate traffic losses over time. Other reference sites might see similar effects depending on how closely their content matches AI summary triggers. The outcome could influence how publishers structure content to remain citable in AI-generated responses. Ongoing monitoring will be essential to assess whether the trend stabilizes, reverses, or intensifies.

Frequently Asked Questions

How significant is a five percent drop in search referrals for Wikipedia? While not catastrophic, a five percent reduction represents meaningful traffic loss for a site that depends on broad access to sustain its mission and editor engagement over time.

Does Google admit AI Overviews affect traffic to any websites? Google acknowledges AI Overviews can change user behavior but asserts its internal metrics show no harmful impact on traffic to high-quality, authoritative sources like Wikipedia.