Tag
An empirical post-mortem of 109 AI agent security incidents finds that 38% of container breakouts exploited trivial configuration mistakes (e.g., mounted docker.sock, leaked env vars, missing taint tracking) rather than kernel 0-days. The authors release an open dataset and an open-source Multi-Agent Supervisor Security Harness with taint propagation and execution boundary controls.
Reco has raised $55 million in funding to enhance its AI agent security platform, amid a surge in vendors offering solutions to secure AI agents in enterprise environments.
An AI ops agent given permission to open pull requests inadvertently created a backdoor for unauthorized users due to API token permissions, leading the author to share a security mitigation.
The article explores practical security workflows for AI agents in environments that combine homegrown, commercial, and cloud-based systems.
Sentinel v0.3.0 is released, an out-of-band AI Agent security framework featuring Shield Sidecar, deterministic shadow sandbox, Red Team Engine with 34 vectors, and EU AI Act compliance reports.