Scaling AI Safety for a Multi-Agent World

Schmidt Sciences

Eligibility

  • Eligible PIs

Description

The call is focused on the emergence of complex ecosystems of AI deployed by different actors across shared digital infrastructure. The research agenda includes broadly: Sandboxes and testbeds, science of agent networks, strengthening infrastructure, and multi-agent oversight.

Learn more about this program

Keywords:
AI, AI safety, multi-agent systems, scalable AI, simulation environments, governance, system-level safety, population-level risk, collective behavior, collective intelligence

Application due:
08/09/26

Award amount:
$300,000 - $1,000,000