Wikipedia Faces Traffic Decline in AI Era
Wikipedia, the world's largest online encyclopedia, is experiencing a significant decline in human traffic as artificial intelligence reshapes how people search for information. According to the Wikimedia Foundation, human pageviews dropped by approximately 8% year-over-year from March to August 2025, marking a concerning trend for the volunteer-driven platform.
The AI Search Revolution
The decline comes as search engines increasingly use generative AI to provide direct answers rather than linking users to external websites. 'When users find what they want from an AI, they don't visit the source, or continue with that session,' explains Marshall Miller, Senior Director of Product at the Wikimedia Foundation, in an official blog post. This behavior results in a loss of traffic for original content sources like Wikipedia.
Google's AI Overviews feature, which provides AI-generated summaries above traditional search results, has been particularly impactful. A TechCrunch report notes that users are half as likely to click traditional search results when AI summaries appear, fundamentally changing search behavior.
Bot Detection Reveals True Picture
The traffic decline became apparent after Wikimedia updated its bot detection systems in May 2025. The foundation had observed unusually high traffic from Brazil that initially appeared to be human users. After investigation, much of this traffic was reclassified as bots designed to evade detection. 'Our algorithms classify traffic from humans and bots, and we've had to update our detection systems to account for more sophisticated bot activity,' Miller stated.
This reclassification revealed the true extent of the human traffic decline, which had been masked by bot activity. The 8% drop represents a significant shift in how people access information in the age of AI.
Irony of AI Dependence
The situation presents a paradox: while AI systems are reducing traffic to Wikipedia, they heavily depend on Wikipedia's content for training. 'Wikipedia's data sets are used to train almost all large language models, but they may not always link to our sites,' Miller noted. According to Observer, Wikipedia, YouTube, and Reddit account for about 15% of AI-generated content, yet these platforms see declining direct traffic.
Broader Industry Impact
Wikipedia's experience reflects a broader trend affecting content publishers worldwide. Industry analysis shows that AI-powered search can reduce website clicks by 18-70%, particularly impacting informational, health, and finance websites. This represents a fundamental shift from Google's traditional role as a gateway to websites to becoming an information destination itself.
'This behavior is not unique to Wikipedia - other platforms and publishers are observing the same,' Miller emphasized, highlighting the widespread nature of this challenge.
Wikipedia's Response
In response to the traffic decline, Wikipedia is taking several strategic approaches. The foundation is working on new attribution frameworks to ensure proper credit for content used by AI systems. They're also establishing two new reader teams focused on Reader Growth and Reader Experience to improve how users access Wikipedia content.
Additionally, Wikipedia is expanding its presence on platforms like TikTok and Roblox to reach younger audiences who increasingly prefer social video over traditional web browsing. 'We're working on a framework for attribution among other improvements,' Miller confirmed.
Quality Concerns and Future Outlook
The traffic decline raises concerns about information quality and verification. 'The problem is that the quality of AI search results is often not better than a traditional search engine,' notes technology journalist Ashwin. 'Bots may spit out a summary based on low-effort posts, inaccurate content, or fabricated posts on social media.'
As AI continues to evolve, the challenge for platforms like Wikipedia will be balancing the benefits of new technology with the need to sustain the human-curated knowledge that makes their content valuable in the first place. The foundation continues to encourage users to visit original sources and support the volunteer-driven model that has made Wikipedia one of the world's most trusted information resources for over two decades.