Wikipedia to License Content to AI Companies
▼ Summary
– The Wikimedia Foundation announced new licensing deals with Microsoft, Meta, Amazon, Perplexity, and Mistral AI to charge them for using Wikipedia content to train AI models.
– These deals expand the Wikimedia Enterprise program, a commercial service that sells high-speed, high-volume API access to Wikipedia’s articles.
– The financial terms were not disclosed, but the revenue helps offset infrastructure costs for the nonprofit foundation.
– Most major AI developers have now joined the program, including Google which signed on in 2022, moving from using free, scraped content to a paid commercial platform.
– According to a foundation official, these tech companies recognize Wikipedia as a critical resource and see the need to financially support its sustainability.
The Wikimedia Foundation has secured new licensing agreements with several major technology firms, allowing them to formally use Wikipedia’s vast content for training artificial intelligence systems. This move formalizes a previously informal practice and establishes a revenue stream to support the nonprofit’s operations. The new partners include Microsoft, Meta, Amazon, Perplexity, and Mistral AI, joining Google and other smaller companies that had already signed on. These deals are managed through the foundation’s commercial arm, Wikimedia Enterprise, which provides high-speed, high-volume API access to Wikipedia’s 65 million articles.
For years, AI developers have freely scraped Wikipedia’s text to train large language models, which power popular assistants like ChatGPT and Microsoft Copilot. The new licensing program asks these commercial entities to pay for reliable, structured access. While the specific financial terms remain confidential, the revenue is intended to help offset the substantial infrastructure costs required to keep Wikipedia running. The encyclopedia relies primarily on small public donations, even as its content has become a foundational element of the AI industry.
Lane Becker, who leads Wikimedia Enterprise, emphasized the importance of this shift. He noted that Wikipedia is an indispensable resource for tech companies and that they must find ways to support it financially. The challenge was designing a commercial service compelling enough for these firms to transition from free, public APIs to a paid platform. According to Becker, the major partners now recognize their responsibility in helping to sustain Wikipedia’s long-term viability. This program represents a strategic effort to ensure the free knowledge project can continue its mission while its data fuels the next generation of commercial AI products.
(Source: Ars Technica)
