Scaling Trust is ARIA’s programme under the Trust Everything, Everywhere opportunity space, backed by about £49.8m. It funds open‑source tooling and fundamental research to let AI agents and automated services securely discover each other, capture requirements, negotiate terms, generate protocols, and verify compliance in adversarial digital and cyber‑physical settings. The programme aims to collapse Coasean transaction costs by making secure multi‑agent coordination cheap and reliable, unlocking new markets and enabling AI systems to act on users’ behalf with verifiable safety and security guarantees.
Endorsements made here support Advanced Research + Invention Agency (ARIA).
Scaling Trust is ARIA’s programme under the Trust Everything, Everywhere opportunity space, backed by about £49.8m. It funds open‑source tooling and fundamental research to let AI agents and automated services securely discover each other, capture requirements, negotiate terms, generate protocols, and verify compliance in adversarial digital and cyber‑physical settings. The programme aims to collapse Coasean transaction costs by making secure multi‑agent coordination cheap and reliable, unlocking new markets and enabling AI systems to act on users’ behalf with verifiable safety and security guarantees.
Endorsements made here support Advanced Research + Invention Agency (ARIA).
Project Details
Updated 05/18/26 · By grantmaking.aiScaling Trust is a large‑scale, approximately £49.8m research and development programme run by ARIA within the Trust Everything, Everywhere opportunity space. Its mission is to build the tools, protocols, and security foundations that allow AI agents to safely coordinate, negotiate, and verify agreements with each other and with human stakeholders, even in adversarial or untrusted digital and cyber‑physical environments.
The programme funds two main pillars of work: (1) open‑source tooling – modular agents and components for secure requirement capture, negotiation, protocol generation, and verification; and (2) fundamental research that upgrades today’s largely empirical AI security practices into theory‑driven guarantees for multi‑agent systems. Projects span short‑term tooling grants and longer research centre‑style efforts, with strong emphasis on reusability and participation in a future Scaling Trust Arena where different tools interoperate.
By dramatically reducing the costs and frictions of establishing and enforcing trust between autonomous systems, Scaling Trust aims to unlock new classes of cyber‑physical markets and enable AI agents to act on behalf of users with measurable trustworthiness and robust security properties.
Theory of Impact
Updated 05/18/26 · By grantmaking.aiScaling Trust starts from the premise that as AI agents proliferate, the bottleneck will be reliable coordination and enforcement of agreements in untrusted environments, not raw model capability. Its theory of change is that providing shared, open‑source tooling and rigorous security and verification frameworks for multi‑agent interactions will drastically lower the transaction costs of coordination, making it easy and safe for AI agents to negotiate and carry out tasks for users. If successful, this infrastructure will enable high‑impact deployments of agentic AI in cyber‑physical systems while preserving security, alignment with user preferences, and robust auditability.
People– no linked people
Updated 05/18/26 · By grantmaking.aiFunding Details
- Feb 10, 2026
- -
- -
- $49,800,000
- -
- -
- -
- -
- -
- -
Discussion
No comments yet. Be the first to share your thoughts.