{
  "id": 15942,
  "url": "https://www.schneier.com/blog/archives/2026/08/the-openai-hack-shows-the-genie-is-out-of-the-bottle.html",
  "title": "The OpenAI Hack Shows the Genie Is Out of the Bottle",
  "summary": "This essay originally appeared in Foreign Policy . Earlier this month, two of OpenAI’s models broke out of their containment sandbox and attacked another AI company. The story is kind of wild . OpenAI was running security tests on two of its models: GPT-5.6 Sol and an unreleased model that is almost certainly GPT-6. In particular, it was running the ExploitGym benchmark, which measures how good a model is at turning security vulnerabilities into working exploits: basically, offensive cyberattack",
  "authors": "Bruce Schneier",
  "category": "org",
  "topics": "regulation",
  "orgs": "openai",
  "regions": null,
  "published_at": "2026-08-03T10:47:47.000Z",
  "fetched_at": "2026-08-04T05:10:21.797Z",
  "source_slug": "x-bruce-schneier-schneier-on-security",
  "source_name": "Bruce Schneier — Schneier on Security",
  "source_homepage": "https://www.schneier.com",
  "ethics_ai_record_url": "https://ethics.ai/record/15942",
  "original_url": "https://www.schneier.com/blog/archives/2026/08/the-openai-hack-shows-the-genie-is-out-of-the-bottle.html",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}