Current AI secures $400M for public AI infrastructure

Current AI has secured about $400 million in committed funding to build AI infrastructure intended for public access rather than private control.

What happened

The nonprofit initiative is backed by a mix of governments, foundations and technology companies. Its funding includes an initial $100 million commitment from the French government, alongside support from organisations including the Ford Foundation, MacArthur Foundation, Google DeepMind and Salesforce.

Current AI is working on projects such as an offline device supporting 22 Indian languages, public-interest AI grants, an open-source conversational model and a collaboration with Japan’s Sakana AI.

Why it matters

Most advanced AI infrastructure is controlled by a small number of companies with the capital to fund large models, specialised chips and data centres. That concentrates access, pricing power and technical decision-making.

Current AI is testing a different model: shared infrastructure that researchers, public institutions and underserved communities can use without depending entirely on commercial platforms.

The bigger picture

The project reflects a wider push for AI public infrastructure. Governments increasingly view access to models, compute and datasets as strategic capacity, similar to broadband, cloud services or research facilities.

The hard part will be sustainability. Public-interest AI still needs large amounts of compute, technical talent and long-term funding. The initiative’s importance will depend on whether it can build useful systems that remain competitive after the initial commitments are spent.