Fund, Multi-Agent Safety

• Support foundational scientific research to understand, evaluate, and control risks from ecosystems of interacting AI agents deployed by multiple actors.
• Develop methods to analyze emergent collective dynamics, build trustworthy agent infrastructure, and create monitoring approaches.
• Address system-level safety gaps including collusion, conflict, destabilizing dynamics, emergent agency, and security vulnerabilities.

This role directly supports foundational research to prevent catastrophic risks from advanced AI systems, ensuring safe and beneficial development of multi-agent AI ecosystems.

• Expertise in AI safety, multi-agent systems, or related fields.
• Experience in research funding, grant-making, or program management.
• Strong analytical and strategic thinking skills.

$10m fund available for research projects. Additional benefits not specified.

The Cooperative AI Foundation supports research and initiatives to ensure AI systems are cooperative and aligned with human values, focusing on safety and governance.