Teleport adds LLM Proxy and Delegated Identity to secure AI agent actions and access

Teleport announced two features in its Agentic Identity Framework—LLM Proxy and Delegated Identity—that give AI agents scoped, auditable, least-privilege access to inference endpoints and infrastructure. LLM Proxy inspects and logs agent requests/responses and enforces which agents may access which inference endpoints, while Delegated Identity issues short-lived, scoped permissions to agents to limit blast radius and enable auditability.