Wikipedia asks AI firms to stop scraping and pay up

The Wikimedia Foundation has called on AI companies to stop freely scraping Wikipedia's content and instead use a paid service, arguing that the future of both the free platform and reliable AI depends on it.

In an official announcement, the foundation stated that human-created knowledge, like that on Wikipedia, is an irreplaceable backbone for AI. It warned that AI models cannot exist without continually updated information curated by humans and risk becoming less accurate without it.

The nonprofit, which supports Wikipedia, is urging AI developers to use its commercial platform, Wikimedia Enterprise. Wikimedia says that this paid-for service allows for large-scale, reliable access to Wikipedia's data without overloading its servers. The foundation frames this as a necessary step for AI companies to sustainably support the human volunteer efforts that create their training data.

Alongside financial support, the foundation is also requesting that AI systems provide clear attribution, crediting Wikipedia when its information is used. It argues that this transparency helps maintain trust and encourages people to visit and contribute to the site directly.

The announcement positions Wikipedia's human-driven, transparent model as a critical antidote to AI systems that can sometimes produce inaccurate or "hallucinated" information, says the foundation. As Wikipedia approaches its 25th anniversary in 2026, the foundation further states that sustaining its model is essential for ensuring the internet continues to provide free, reliable knowledge.