Researchers call OpenAI's security practices negligent, undermining its AI claims
BlancheMinerva · x · 2026-09-14
Quoting Joshua Saxe's point that OpenAI has been "obviously negligent in its security practices," BlancheMinerva says the prevailing attitude is: if OpenAI can't be bothered to follow real security practices for years, why believe its claims about AI capabilities against good security? The thread highlights a credibility gap between OpenAI's security culture and its safety-related claims.
More from Companies & People
- a16z: AI agents' task length doubles every 7 months as Databricks bets on agents' system of record — _AustinCalvert_ · 2026-10-28
- NeurIPS activates Sydney & Atlanta waitlists, releases more spots Oct 30 — NeurIPSConf · 2026-10-28
- AI policy researcher Janet Egan joins IFP as director of emerging tech — RishiBommasani · 2026-10-28
- Artificial Analysis hosts "Independent analysis of AI" event in San Francisco — echen · 2026-10-28
- Apple cuts 200+ jobs across Siri, Vision Pro and AI teams under new CEO Ternus — CtrlAltDwayne · 2026-10-28
- Big company's Head of AI: 10% of employees drive 90% of AI productivity gains — _AustinCalvert_ · 2026-10-28