Wikipedia reports 8% fall in human traffic amid rise of AI search and social video
Google has, however, disputed claims that its AI search summaries reduce traffic to publishers.
ADVERTISEMENT
Wikipedia, long regarded as one of the internet’s last bastions of credible, community-driven knowledge, is now seeing a noticeable decline in readership. The Wikimedia Foundation has reported an 8 percent year-on-year fall in human pageviews, attributing the drop partly to the growing influence of AI search tools and social video platforms.
In a blog post, Marshall Miller of the Wikimedia Foundation explained that recent updates to Wikipedia’s bot detection systems revealed that much of the unusually high traffic observed during May and June came from bots designed to evade detection. Once the data was corrected, the true extent of the fall in human visits became clear.
Miller said the decline is linked to changing user behaviour, as generative AI and social media alter how people access information. Search engines are increasingly using AI-generated summaries that answer questions directly on results pages instead of linking to external websites such as Wikipedia. At the same time, younger audiences are turning to social video platforms for quick, visually-driven explanations rather than traditional web searches.
Google has, however, disputed claims that its AI search summaries reduce traffic to publishers.
Despite the decline, Miller maintained that Wikipedia’s role in the digital knowledge ecosystem remains crucial. He said that while fewer users are visiting the site directly, information sourced from Wikipedia continues to reach people indirectly through AI tools and other digital platforms. The organisation had even trialled AI-generated summaries on its own platform but paused the initiative following concerns from volunteer editors.
Miller acknowledged that the shift in browsing habits presents long-term challenges. With fewer visitors, the pool of volunteer editors and individual donors could shrink, potentially impacting the quality and sustainability of the project.
To address this, Miller said AI, search, and social media companies using Wikipedia’s content must encourage users to visit the site directly. The foundation itself is working on a new content attribution framework to ensure Wikipedia remains properly credited across the web. Two internal teams have also been tasked with developing strategies to reach new readers and strengthen community engagement.
Miller urged readers to take an active role in supporting trustworthy information online. He encouraged people to verify sources, click through to original material, and remind others that the knowledge powering generative AI models originates from human contributors who deserve recognition and support.
As Miller put it, the challenge for Wikipedia is not just preserving traffic, but ensuring the continued integrity of human-curated knowledge in an internet increasingly shaped by machines.
