Monday, December 8, 2025

-title-generator Wikipedia sees decline in human pageviews, says AI is to blame

Wikipedia ‍Reports Traffic Drop, Cites⁣ Impact of AI Search Summaries

San francisco, CA – October 20, 2025 – Wikipedia is experiencing a decline in human pageviews, attributing the ⁢drop to the increasing use of AI-powered search ⁢summaries, chatbots, and social⁤ media ⁢content. According to a new report from the⁣ Wikimedia Foundation, traffic to Wikipedia sites fell by 8%⁤ in 2024 compared to the previous year.

Wikimedia believes this​ decrease stems from the fact that its data is extensively used to train large language models (LLMs), yet ⁤these models don’t consistently link back to​ Wikipedia as a source. The foundation notes this ​trend isn’t isolated ‌to Wikipedia, with other⁢ publishers also observing similar impacts.

The findings challenge previous statements from Google,⁤ which has denied that AI in search results reduces traffic to websites,despite reporting that ⁢ Google AI Overviews has over⁢ 2 billion monthly users.

The‌ core issue, Wikimedia explains, is that users satisfied by AI-generated responses are less likely to visit the original source⁢ material.”When a user finds what they want from ⁤an AI, they don’t visit the source, or continue with that session,” the report indicates.

Wikipedia itself briefly experimented with AI-generated summaries⁤ in June, but faced criticism from its editors, who deemed the practise unethical, as reported by TechCrunch.

While acknowledging the convenience of AI search for tasks like shopping and product ⁢details gathering, the report also raises concerns about the quality and accuracy of AI-generated content.⁣ The author notes that AI summaries ‍can be based ⁢on ⁢”low-effort posts, inaccurate⁣ content,​ fabricated posts on social media,” and warns against ⁢blindly accepting AI outputs, referring to inaccuracies as​ “hallucinations” – a term they find misleading, as AI models simply reproduce ​information ‌from their training data.

Wikimedia is actively working to address the issue, launching new teams focused ⁢on reader growth and experience, and developing a framework for ⁤proper attribution to source websites. The foundation is encouraging AI users ‌to visit the original source material to ensure accuracy and support content creators.

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.