{
 "license": "CC-BY-4.0",
 "attribution": "Fide AI, Agentic Cyber Explorer",
 "url": "https://agentic-cyber-explorer.pages.dev/events/openai-expanded-misalignment-monitoring-2026/",
 "asOf": "2026-10-30",
 "id": "openai-expanded-misalignment-monitoring-2026",
 "date": "2026-10-29",
 "datePrecision": "day",
 "title": "OpenAI describes expanded agent monitoring, with workload-specific coverage and response limits",
 "lane": "defense",
 "kind": "eval-report",
 "summary": "OpenAI describes expanded misalignment monitoring across training, evaluations, internal use and selected external deployments, using activation probes, model reviewers, investigative agents and telemetry. Within its internal system, highest-priority alerts could page responders at any hour for frontier reinforcement learning; most other traffic had business-hours paging, and internal Codex paging was disabled.",
 "whyItMatters": "It makes the distinction between monitoring coverage, escalation and actual intervention explicit.",
 "actors": [
  "openai"
 ],
 "topics": [
  "monitoring-and-control",
  "sandbox-containment",
  "autonomous-defense",
  "eval-validity"
 ],
 "atlas": [
  "monitor",
  "model",
  "sandbox"
 ],
 "artifacts": [
  "gpt-6-astra"
 ],
 "sources": [
  {
   "url": "https://openai.com/index/how-we-detect-frontier-model-misalignment/",
   "publisher": "OpenAI",
   "title": "How we detect frontier model misalignment",
   "date": "2026-10-29",
   "type": "primary",
   "accessed": "2026-10-30",
   "shared": true
  }
 ],
 "keyFacts": [
  {
   "fact": "Replaying the Hugging Face incident produced an alert about 30 minutes after sandbox escape and over 30 hours before the historical breach. This was a retrospective simulation, not prevention of that incident.",
   "locator": "Introduction"
  },
  {
   "fact": "Frontier RL and internal Codex traffic go directly to the model-based reviewer, skipping activation probes. For these workloads, OpenAI was moving to no more than 30 minutes between checks; a broader planned pause would act if responders had not ruled out a highest-priority alert within 30 minutes of paging.",
   "locator": "Layer 0; Layer 1; Human response"
  },
  {
   "fact": "External monitoring does not routinely escalate to OpenAI human reviewers. For covered models, API monitoring applies to tool-using Responses requests; automatic stopping requires supported conversation tracking, and Chat Completions is excluded. OpenAI says one fix from the July METR/Redwood exercise remained in progress.",
   "locator": "Safeguarding external usage of Astra; Red-teaming"
  }
 ],
 "significance": 4,
 "fideQuestions": [
  "FID-074",
  "FID-075",
  "FID-076"
 ],
 "methods": [
  "ai-monitoring"
 ],
 "review": "assistant-drafted",
 "addedOn": "2026-10-30",
 "revisions": []
}