Wikipedia Calls on AI Firms to Use Paid API, Stop Scraping

Wikipedia, the world’s largest free encyclopedia, has long relied on volunteer editors and a generous user base to keep its millions of articles up to date. Its content is freely available under a Creative Commons license, making it a natural source of training data for machine‑learning models. In recent years, however, a surge in AI development has led to large companies crawling the site at high speed to harvest facts, phrasing, and even entire pages. This scraping puts additional strain on Wikipedia’s servers, risks violating its terms of service, and threatens the sustainability of the volunteer‑driven model.

To address these concerns, Wikipedia’s board issued a public call in a TechCrunch report, urging AI firms to stop scraping and instead subscribe to its newly introduced paid API. The API delivers structured, indexed content in a way that reduces server load and gives developers a clear licensing path. While the free content remains openly accessible, the paid tier offers higher request limits, priority support, and legal clarity. Wikipedia hopes that the revenue generated through the API will help fund infrastructure costs, pay for moderation tools, and support its volunteer community.

For AI developers, the shift presents a cost–benefit dilemma. On one hand, the API’s fees can add up for companies that rely heavily on Wikipedia for facts, code snippets, or historical data. On the other, continued scraping exposes them to potential legal challenges, IP disputes, and reputational risk. Some firms are already exploring alternative data sources or negotiating bulk licensing agreements. The broader debate underscores how the growth of AI is reshaping the economics of open knowledge, forcing a conversation about fair compensation, data stewardship, and the future of freely available information.

Key takeaway: AI developers must weigh the costs of licensing against the risks of unsanctioned scraping, as Wikipedia seeks to monetize its freely available knowledge.