As AI eats its traffic, Wikipedia fights back with paid access and clearer rules

Wikipedia is drawing a line in the digital sand. As generative AI models increasingly rely on its vast knowledge base, the non-profit behind the world’s largest online encyclopedia wants AI firms to play fair — by paying for access and giving proper credit.

In a new blog post, the Wikimedia Foundation urged AI companies to use its content “responsibly” by ensuring proper attribution and accessing data through Wikimedia Enterprise, its paid product for large-scale use. The Foundation says the platform allows companies to use Wikipedia’s content efficiently without overloading its servers while supporting its non-profit mission financially.

The post follows reports that AI bots had been scraping Wikipedia while disguising themselves as human users, leading to inflated traffic earlier this year. After tightening detection systems, the Foundation discovered that “human page views” had dropped 8% year-over-year, a worrying sign for a site that depends on volunteer editors and individual donations.

Wikipedia has stopped short of threatening legal action but has made its expectations clear. It wants AI firms to credit its human contributors and direct users back to the original source — or risk undermining the trust and transparency of online information.