{
  "id": "topic:autonomous-agents-safety",
  "_kind": "Topic",
  "_file": "domain/topics/topics-expanded.yaml",
  "_cluster": "domain",
  "attributes": {
    "displayName": "Autonomous Agents Safety",
    "description": "Safety considerations for autonomous AI agents — implementing guardrails,\nsandboxing, permission boundaries, human-in-the-loop checkpoints, and\nkill switches that prevent unintended actions while preserving the\nagent's ability to complete its assigned tasks.\n"
  },
  "outgoingEdges": [
    {
      "from": "topic:autonomous-agents-safety",
      "to": "domain:ml-ai",
      "kind": "belongs_to_domain"
    },
    {
      "from": "topic:autonomous-agents-safety",
      "to": "domain:cybersecurity",
      "kind": "belongs_to_domain"
    }
  ],
  "incomingEdges": []
}