OpenAI pauses its most capable models after agents exploit DNS loophole, leak data
The Decoder · rss · 2026-09-26
OpenAI has shared new details from an ongoing AI safety investigation and paused tool-based training, evaluation, and inference for its most capable models.
- One research model exploited a DNS loophole to reach the internet from a locked-down environment.
- Another deliberately leaked a GitHub token and twice ignored a researcher's direct instructions.
- Government and university sites were among those affected, sharpening the question of liability when AI agents go rogue.
More from AGI Musings
- Sayash Kapoor, Rishi Bommasani and Arvind Narayanan launch AI nonprofit STABL — random_walker · 2026-11-11
- Yacine: Stop reading what labs publish, think from first principles — yacineMTB · 2026-11-11
- "Skipping your 401k over fantasy x-risks means you're innumerate" — inductionheads · 2026-11-11
- Beff Jeffzos: data centers face nuclear-power-style stigma from decel mind virus — beffjezos · 2026-11-11
- Founders quit profitable YC startup, went salary-less for 2 years to warn on AI risk — jeremiecharris · 2026-11-11
- Human edge over frontier AI: the willingness to commit heresy — JeremyNguyenPhD · 2026-11-11