Ex-OpenAI Researcher Criticizes AI Safety Culture as Superficial
Miles_Brundage · x · 2026-08-07
Former OpenAI policy researcher Miles Brundage emphasizes that while publishing technical reports is important, a company's internal safety culture and processes are equally—if not more—crucial.
He also questioned the recent safety evaluation of a model, noting that after a significant cybersecurity incident, the team seemingly resumed business as usual just two days later. He views this as a sign that safety reviews might be becoming merely performative.
Related event: Former OpenAI Policy Lead Criticizes Internal AI Safety Culture(2 posts)→
More from Safety
- Genome Language Models Can Make Viable Phages, But Bioweapon Barriers Remain — teortaxesTex · 2026-08-07
- Claude Tries to Merge Malicious Code: Is Persona Alignment Just a Fragile Shell? — NathanpmYoung · 2026-08-07
- Artificial Intelligence Used to Design New Viruses, BBC Reports — CaRDiaK · 2026-08-07
- USA Today Owner Gannett Partners with Palantir to De-anonymize Reader Data — SatelliteNetSec · 2026-08-07
- AI Model Sandbox Escapes Will Soon Become Undetectable — jachiam0 · 2026-08-07
- MIT Paper: Undetectable Covert Conversations Between AI Agents via Steganography — geoffreyirving · 2026-08-07