The Wikimedia Foundation\’s Wikimedia Enterprise unit announced licensing agreements with Microsoft, Meta, Amazon, Perplexity, and Mistral AI to monetize Wikipedia content used to train AI models. This marks a shift from free access toward paid partnerships for the data that underpins popular AI assistants.
\n
Under the new arrangements, these firms will license API access to Wikipedia\’s 65 million articles at higher speeds and volumes than the free public APIs provide, with terms that the foundation has not disclosed. Google, which signed a deal with Wikimedia Enterprise in 2022, remains among the platform\’s partners, alongside smaller players like Ecosia, Nomic, Pleias, ProRata, and Reef Media.
\n
The revenue is intended to offset infrastructure costs for Wikimedia, a nonprofit that relies on donations and faces growing demand for its content as it becomes embedded in AI training data and tooling. Lane Becker, president of Wikimedia Enterprise, told Reuters that major tech partners view sustaining Wikipedia\’s work as a shared responsibility.
\n
Wikimedia Enterprise emphasizes faster, more scalable access to Wikipedia content, enabling AI developers to incorporate Wikipedia data into training pipelines more reliably than relying on public scrapes. While the precise financial terms were not disclosed, the deals represent a broad shift toward revenue-backed support for the project.
\n
As licensing expands, the Wikimedia Foundation expects the collaboration to help ensure continued quality and maintenance of its vast encyclopedia, while reflecting a broader industry trend of monetizing access to high-value public knowledge for AI systems.
