The Wikimedia Foundation, which oversees Wikipedia, has issued a public appeal to developers of artificial-intelligence systems to stop scraping its public pages and instead access its content via the paid Wikimedia Enterprise API.
In its announcement, the foundation stressed that the Enterprise product is designed to handle large volumes of requests without undermining Wikipedia’s volunteer-based infrastructure and that revenues from this service support its non-profit mission to keep knowledge free and accurate.
The move reflects a growing tension between open-knowledge platforms and large AI models, which frequently train on publicly available data at scale without direct compensation to the content creators.
The foundation noted a decline in human page views of about 8 percent year-over-year amid increased bot traffic, which it attributes in part to AI bots scraping the site.
From Volunteer Archives To Commercial Licences
Wikipedia’s platform relies on a global community of volunteer editors and donations rather than advertising revenue.
The foundation argues that unchecked scraping by AI systems threatens the feedback loop of human contribution, page visits and donor support.
“With fewer visits, fewer volunteers may grow and enrich the content, and fewer individual donors may support this work,” the blog post stated.
By offering Wikimedia Enterprise, the foundation provides a route for companies to access its content under terms that include attribution, metadata disclaimers and stability.
The paid API aims to enable high-volume usage by AI, search and knowledge-graph platforms in a way that does not “severely tax Wikipedia’s servers.”
While the foundation stopped short of legal threats, its expectation is clear: AI firms that rely on Wikipedia’s human-generated content should contribute back through licensed access rather than free scraping.
Does This Matters For AI And Knowledge Ecosystems
This development signals a broader shift in how content platforms view data extraction in the era of generative AI. For AI system builders, it raises questions about sourcing, attribution and cost of large-scale training data sets.
Platforms with substantial human labour behind them are increasingly asserting rights to be compensated. For Wikipedia and similar knowledge repositories, this may become a precedent.
The foundation is taking a proactive step to ensure that as AI models leverage its content, the ecosystem that sustains free knowledge is not undermined. It may also influence regulatory and ethical standards around training data usage in AI.
As Wikipedia transitions from offering purely open access to providing enterprise-grade licensed data, the question becomes how AI companies will respond.
Will they pay for access and attribute correctly, or will scraping continue unchecked? The outcome will shape how knowledge is sourced, trained and monetised in the next phase of AI.
