The Wikimedia Foundation has announced new licensing deals with major AI companies including Microsoft, Meta, and Amazon to provide paid access to Wikipedia content. These partnerships aim to offset rising infrastructure costs caused by AI scraping. The deals mark a shift from unauthorized data use to commercial API access through Wikimedia Enterprise.
On January 15, 2026, the Wikimedia Foundation revealed partnerships with AI developers such as Microsoft, Meta, Amazon, Perplexity, and Mistral AI as part of Wikipedia's 25th anniversary celebrations. These companies, previously known for scraping Wikipedia's vast repository of 65 million articles without permission, have now joined the nonprofit's commercial subsidiary, Wikimedia Enterprise. The program offers high-throughput APIs for faster, higher-volume access to Wikipedia and related projects like Wikivoyage, Wikibooks, and Wikiquote, helping to sustain the organization's operations amid surging costs.
The initiative addresses a growing financial strain on the foundation, which relies primarily on small public donations. Last year, Wikimedia raised alarms about an existential threat from reduced website traffic due to large language models (LLMs) and AI chatbots summarizing content without directing users to the source. In April 2025, bandwidth for downloading multimedia content increased by 50 percent since January 2024, with bots accounting for 65 percent of the most expensive infrastructure requests despite comprising only 35 percent of total pageviews. By October 2025, human traffic had declined about 8 percent year-over-year after improved bot-detection measures revealed many 'visitors' were automated scrapers.
This traffic drop disrupts Wikipedia's traditional feedback loop, where readers become editors or donors, enhancing content quality. Meanwhile, AI firms use the human-curated data to power tools like Microsoft Copilot and OpenAI's ChatGPT. Lane Becker, president of Wikimedia Enterprise, emphasized the importance of financial support: “Wikipedia is a critical component of these tech companies’ work that they need to figure out how to support financially... all our Big Tech partners really see the need for them to commit to sustaining Wikipedia's work.”
Wikipedia founder Jimmy Wales supports AI training on the data but insists on compensation: “I’m very happy personally that AI models are training on Wikipedia data because it’s human curated... You should probably chip in and pay for your fair share of the cost that you’re putting on us.” The new deals join earlier ones, such as Google's 2022 agreement, though financial terms remain undisclosed. The foundation has faced internal resistance to its own AI experiments, pausing a pilot for AI-generated summaries in June after editor backlash.
