Wikipedia Tells AI Firms to Stop Scraping and Pay

Wikipedia has a message: its human core is crucial in the age of artificial intelligence.

In an announcement from the Wikimedia Foundation, it says every AI model, no matter how advanced, depends on the vast, carefully documented work of human editors. That is why the Wikimedia Foundation says the role of Wikipedia — the world’s largest collaborative knowledge project — has never been more critical.

“Many people do not recognize that there is something distinctly familiar behind all that AI-generated content: knowledge curated, debated, and documented by humans,” the Foundation writes.

Why AI needs Wikipedia

Generative AI systems have learned vast amounts from open, nonprofit datasets — Wikipedia among them. Studies show that when AI developers omit Wikipedia from their training data, their models produce less accurate, less diverse, and less verifiable outputs.

This dependence has prompted calls for ethical collaboration. The Wikimedia Foundation is asking AI companies to support Wikipedia in two key ways: through attribution and financial support.

Attribution ensures that AI tools credit the human sources they rely on, reinforcing the feedback loop that keeps information accurate and trustworthy. Financial support, meanwhile, helps maintain Wikipedia’s infrastructure through the Wikimedia Enterprise program — a paid access system that allows commercial users to sustainably draw from Wikipedia’s data without overburdening its servers.

By contributing to this ecosystem, AI companies protect their own future as well. Without Wikipedia’s constant human updates, AI systems risk spreading misinformation and losing public trust.

The value of human editors

Hank Green, a well-known science communicator, recently asked whether AI might “end up eating itself.” The concern is that without new, human-created information, AI tools will degrade over time. Wikipedia helps prevent that scenario.

Its volunteer editors engage in constant discussion, verification, and consensus-building. Unlike AI, they can debate context, uncover forgotten details in archives, and document underreported regions and cultures. This human effort produces multilingual, balanced, and verifiable content across more than 300 languages — something that artificial systems still struggle to replicate.

Wikipedia’s transparency also sets it apart from algorithm-driven platforms. Every edit, source, and citation can be traced.

“Everyone sees the exact same information on Wikipedia,” the Foundation notes. There are no hidden algorithms personalizing results for profit. Each statement can be verified through cited sources, allowing users to see where knowledge comes from and how it evolves.

The Foundation’s cautious embrace of AI

While AI depends heavily on Wikipedia, the platform is also experimenting with AI assistance — on its own terms. The Wikimedia Foundation’s strategy emphasizes that technology should support human editors, not replace them.

A large portion of volunteers’ time is spent on repetitive maintenance tasks, such as identifying vandalism or cleaning up broken links. AI tools could automate some of these duties, freeing editors to focus on deeper, more complex contributions like writing and verifying content.

Crucially, these tools are being designed and governed by the volunteer community itself. Volunteers create and enforce rules for how AI is used across Wikipedia, ensuring accountability and alignment with the platform’s core values of neutrality and transparency.

The only truth

As the site approaches its 25th anniversary on 15 January 2026, its mission remains unchanged: to provide free, accurate, and human-created knowledge for all.

In an era when AI-generated content floods digital spaces, it reckons the human hands behind Wikipedia may be the most valuable safeguard for truth that the internet has left.

Researchers just answered the question we’ve all been wondering: if AI agents competed directly against human workers on the same jobs, who would win?