Uber open-sources ADR to secure enterprise AI agents
Uber has open-sourced Agentic Detection and Response (ADR), a security framework designed to monitor and defend enterprise AI agents. Deployed internally at Uber, ADR equips organizations with real-time observability, automated security benchmarking, and proactive threat detection tailored to the unique operational risks of autonomous AI workflows and tool-augmented model executions.
As enterprise deployment of autonomous AI agents accelerates, traditional infrastructure security falls short—making agent-specific threat detection frameworks like ADR essential for corporate safety.
* Extends traditional detection and response paradigms to autonomous AI agent workflows and tool calls.
* Combines real-time runtime observability with pre-deployment security benchmarking for AI models.
* Proves its practical viability by already securing high-scale production AI agent deployments at Uber.
DISCOVERED
1h ago
2026-08-04
PUBLISHED
1h ago
2026-08-04
RELEVANCE