Mozilla Data Collective raises funding to expand a governed AI data marketplace
The platform lets dataset contributors choose how their material is licensed, accessed or compensated, giving AI builders a route to multilingual and multimodal data with clear sharing terms.
Data access with contributor-set terms
Mozilla Data Collective connects AI developers looking for training data with organisations and communities that can supply it. Contributors retain ownership and choose whether a dataset is openly licensed, community-governed, compensated or restricted to approved requests; that choice becomes part of how builders discover and access the material.
The model addresses a practical gap in AI development: a dataset’s usefulness depends not only on its contents, but also on whether its provenance and permitted uses are clear. A catalogue spanning multilingual, multicultural and multimodal data gives builders a discovery layer, while access controls let contributors determine when sharing requires review.
Scaling the supply of usable datasets
The company says its catalogue covers more than 450 languages. The new funding is intended to widen language and text coverage across Europe, Africa and South Asia and develop additional ways for organisations to share data with greater control. That makes the investment a bet on broadening both the supply of training material and the rules that make it usable.
Use of funds
Expand multimodal datasets and larger text corpora across EU, African and South Asian languages; develop new ways for organisations to share data with greater control