Skip to content

Schmidt Sciences – Scaling AI Safety for a Multi-Agent World

Home / Funding Opportunity / Schmidt Sciences – Scaling AI Safety for a Multi-Agent World

Schmidt Sciences – Scaling AI Safety for a Multi-Agent World

twitterlinkedin

This is a joint funding call by  Schmidt SciencesGoogle DeepMindAdvanced Research & Invention Agency (ARIA), the Cooperative AI Foundation, and Google.org.

Proposals are sought for foundational scientific research needed to understand, evaluate, and control risks emerging from large-scale ecosystems of interacting AI agents, deployed by multiple actors.

Proposals may target one or more of the four research clusters.  Priority will be given to proposals that target depth rather than breadth.

Research Clusters

  1. Sandboxes and Testbeds – realistic, reproducible multi-agent environments.
  2. Science of Agent Networks – safety-relevant properties of interacting agent populations: how collective capabilities emerge and scale; how networks of agents fail or become volatile; how dangerous population-level properties can be detected.
  3. Strengthening Agent Infrastructure – evaluation and stress-testing of the technical primitives (identify, verifiability, reputation, communication, commitment) on which trustworthy multi-agent interactions will depend.
  4. Multi-Agent Oversight and Control – detection, attribution, security and intervention methods needed to keep deployed agent populations safe at scale.

Two tiers of funding are available:

Tier 1 – up to USD$300,000 – exploratory research projects, pilot studies or focused technical investigations

Tier 2 – USD$300,000 – $1M – more ambitious or collaborative projects.

Projects are to be of one to two years duration.

Proposals are due 8 August 2026.

 

Read more

You may be also interested in ...