Distributed trust at scale
Enterprise AI agent systems are no longer confined to one environment. Your AI ecosystem might span:
- ChatGPT in Azure, authenticated via Entra
- LangChain in AWS, with SPIFFE-based identities
- CrewAI agents on-premises, running with PKCE
- Open-source agents embedded in CI/CD, deployed via GitHub Actions
In this world, trust must be distributed but enforceable.
The agent fabric supports federated identity models, connecting agents across:
- Multiple clouds (Azure, AWS, GCP)
- Multiple IDPs (Okta, Entra, Auth0, Keycloak)
- Different platforms (LLM tools, agent frameworks, container systems)
Just like the identity fabric federates human authentication across providers, the agent fabric federates agent identity across platforms — and enforces consistent policies, scopes, and audit rules.
