In January 2026, the Wikimedia Foundation, parent company of Wikipedia signed strategic partnerships with Amazon, Meta, Microsoft, Mistral AI, and Perplexity to provide structured, paid access to Wikipedia content for Artificial Intelligence(AI) training.
- Marking Wikipedia’s 25th anniversary, these companies will access high-speed data through Wikimedia Enterprise to support the development of Large Language Models (LLMs) and chatbots.
- What? Signed paid partnerships to provide Wikipedia content for AI training.
- Who? Wikimedia Foundation with Amazon, Meta, Microsoft, Mistral AI, Perplexity
- Objective: Support development of LLMs and chatbots
- Programme: Wikimedia Enterprise Programme
Key Details of Partnership:
Content: Wikipedia’s content, over 65 million articles across 300+ languages, is extensively used as training data for AI systems, including chatbots, search engines, and virtual assistants.
Programme: These agreements fall under the Wikimedia Enterprise programme, which offers high-volume, structured Application Programming Interface (API) access specifically designed to meet the requirements of commercial AI developers and technology platforms.
Licensed Access: The move signifies a transition from unauthorised free scraping of Wikipedia data by AI firms to formal licensing arrangements, aimed at ensuring sustainable revenue for the non-profit’s operations.
Expansion of Licensing Partnerships: The new agreements build on earlier deals, including the 2022 partnership with Google, bringing most major technology companies under the Wikimedia Enterprise licensing framework.
No comments yet. Be the first!