It's AI where compute, data, governance, and oversight are distributed between countless parties. Think distributed intelligence, not centralized control.

• It unlocks collaboration between different entities without surrendering control
• It reduces single-point failures & lock-in
• It enables safer, auditable AI systems
• It's the fastest path to achieving AI commons

Most members are distributed between Oxford, Stanford, and Austin.

Multiagent systems, distributed oversight, and anything that makes AI operate better without relying on a central entity.

If you're an organization or an individual looking for a collaboration: collaborations@decentralized-ai.org.

If you're looking for a visiting position: see below.