Blog

Robotics & embodied AI

fleetai-safetyoperationshuman-in-the-loop

Fleet Kill Switch for Autonomous Agents & Robots

Org-wide pause returns BLOCKED on every verify until you resume. One operator action stops agents, robots, and workflows immediately.

May 13, 20266 min read

Incident response for autonomous fleets needs a big red button that works in seconds — not a ticket to disable API keys one service at a time. Sanctum’s org-wide kill switch returns BLOCKED on every verifyAction until an operator resumes.

When to use it

  • Suspected prompt injection across multiple agents
  • Bad model deploy or policy misconfiguration
  • Security exercise or regulated maintenance window

What still works

Audit logging continues. Operators review the queue in the console or mobile PWA. Resume restores normal policy evaluation — no redeploy required.

Compliance evidence · Enterprise

Guides: agentic AI risk · MCP security · runtime authorization · HITL approvals · coding agents · get startedMore: all posts · AI trust layer · open Sanctum Console

Give every agent action a trust boundary.

Start with Connect Agent, keep the SDK path for deeper fleets, and prove exactly what was approved, blocked, or contained.