Google DeepMind launches a $10 million fund for multi-agent AGI safety research

sebkrier · x · 2026-07-26

Google DeepMind, Schmidt Sciences, the Cooperative AI Foundation, ARIA and Google.org are launching a new $10 million fund for research on multi-agent, multi-principal AGI safety.

The initiative shifts attention away from single-agent alignment and centralized control toward problems such as mechanism design, cooperative AI, and coordination in multi-agent settings. Researchers can now apply for support through the linked call for proposals.

Original post →

More from Safety

Safety channel →