Wikimedia urges AI firms to stop scraping its content and to use its paid Enterprise API

The Wikimedia Foundation has formally requested that major artificial intelligence companies stop scraping Wikipedia and instead access its content through the paid Wikimedia Enterprise API. The organization said this model helps sustain Wikipedia as human traffic declines, allowing companies structured, large-scale access while reducing server strain and supporting its nonprofit mission. Companies including Google, OpenAI, Meta, Anthropic, Microsoft, Perplexity, DeepSeek, and xAI have been contacted, though none have publicly commented.

Wikimedia warned that the growing use of AI chatbots like ChatGPT reduces direct visits to Wikipedia, threatening its donation-based funding model. The foundation found that AI bots had inflated traffic earlier this year while human page views fell 8 percent year over year, prompting concerns about lower engagement and fewer volunteer contributions. As more users rely on AI-generated answers instead of visiting the site, donations may decline, a serious issue given its $179 million in annual operating costs funded mainly by public contributions since the site runs without ads.

Along with requesting paid and transparent data access, Wikimedia asked AI companies to credit the editors whose work supports their models. The foundation framed this as an ethical guideline for responsible AI development, avoiding legal threats and aligning with a broader push from content creators demanding compensation when AI systems reuse online material without permission, while promoting fair collaboration.