Wikipedia signs AI training deals with Microsoft, Meta, and Amazon (arstechnica.com)

The Wikimedia Foundation has announced new licensing agreements with major tech companies, including Microsoft, Meta, and Amazon, aimed at monetizing the use of Wikipedia content for training AI models. These partnerships, which also include Perplexity and Mistral AI, signify a shift from companies that previously scraped Wikipedia data without permission to now supporting the platform financially through its Wikimedia Enterprise program. This initiative offers enhanced API access to Wikipedia’s extensive database of 65 million articles, catering to the needs of AI developers while generating revenue to support the nonprofit’s infrastructure costs. This development is significant for the AI/ML community as it establishes a precedent for financial accountability and support from tech giants that rely heavily on open-source content for AI training. By engaging in these agreements, companies like Google and the newly partnered firms recognize Wikipedia's vital role in developing AI technologies. According to Lane Becker, president of Wikimedia Enterprise, this move not only addresses ethical considerations but also ensures that Wikipedia can continue to thrive as a crucial resource in an AI-driven landscape. The details of the financial arrangements have not been disclosed, but this strategic shift indicates growing recognition of the need for sustainable funding in the AI ecosystem.