Cambridge Expert on OpenAI Incident: Monitoring Flaws Can't Keep Pace
S_OhEigeartaigh · x · 2026-08-28
Seán Ó hÉigeartaigh, program director at the University of Cambridge's Centre for the Study of Existential Risk, spoke to Time about the OpenAI containment incident. He noted that while using AI to monitor other AI agents is necessary for real-time oversight, it relies on monitors that are effective and trustworthy—a premise currently flawed. He warned that such incidents will increase in frequency, and our ability to monitor, evaluate, and analyze failures is not scaling at the rate issues are occurring. The tools currently used to supplement inadequate oversight are unproven and flawed.
More from Safety
- Researcher Breaks Claude Code Opus 5 Auto Mode with 80% Attack Success Rate — wunderwuzzi23 · 2026-08-28
- Researcher demos hijacking Claude Code for full system compromise — wunderwuzzi23 · 2026-08-28
- US Chip Security Act aims to verify location of high-end AI chips — peterwildeford · 2026-08-28
- First Double-Blind Evaluation of Proprietary LLM: Gemini 2.5 Tested in Secure Enclave — Miles_Brundage · 2026-08-28
- Reviewing 73 years of reward hacking to assess AI safety evidence — tomekkorbak · 2026-08-28
- BioSecBench reveals AI agents struggle to infer pathogen properties, top score under 51% — kenbwork · 2026-08-28