Fired OpenAI safety researchers dispute misconduct claims
Three fired OpenAI safety researchers dispute allegations of mishandling sensitive information. In an open letter they warn the dismissals create a chilling effect on the company's AI safety culture.
First seen 8 Oct, 20:04 UTC on TechCrunch5 sourcesLast update 11d ago
OpenAIIncident · 8 Oct
Security incident
What the sources say
Linked, never rewritten. Official means the lab itself.
Official0
No word from OpenAI yet.
Press3
TechCrunch· 8 Oct, 20:04Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effectThe Verge· 9 Oct, 09:48OpenAI doubles down on decision to fire three AI safety researchersThe Decoder· 9 Oct, 11:45OpenAI's safety crisis keeps getting worse and the company keeps making it worseHow it unfolded
Every source in the order it appeared. Times in UTC.
8 Oct 20:04TechCrunch Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effect9 Oct 09:48The Verge OpenAI doubles down on decision to fire three AI safety researchers9 Oct 11:45The Decoder OpenAI's safety crisis keeps getting worse and the company keeps making it worse9 Oct 12:21Hacker News Iranian campaign planted fake articles in real U.S. publications using ChatGPT9 Oct 12:31r/OpenAI The OpenAI employees: "We were fired for prioritizing safety over the near-term interests of OpenAI as a corporation."
What the numbers did
Signals for OpenAI from two hours before this story to ten days after: leaderboards, app charts, usage, prices and status.
▼StatusOpenAI reports degraded service: Delayed data in the Costs API▲UsageGPT-6 Astra enters OpenRouter's top 10 models by spend, at #1▼UsageOpenAI's share of OpenRouter spend down 1.5 points in a week, to 26.9%•StatusOpenAI's status page is back to normal after 71 hours▼StatusOpenAI reports degraded service: Degraded performance in GPT-6.1 Sol Work mode•StatusOpenAI's status page is back to normal after 8 hours▼StatusOpenAI reports degraded service: Degraded performance in Codex and ChatGPT Work•StatusOpenAI's status page is back to normal after 3 hours▼StatusOpenAI reports degraded service: Limited availability on gpt-image-2.5-flare and gpt-image-2.5-sunburst•StatusOpenAI's status page is back to normal after 45 minutes▼StatusOpenAI reports degraded service: Delayed Costs data in the Compliance API•StatusOpenAI's status page is back to normal after 42 hours
Keep up with OpenAI
OpenAI on the ScoreIts timelineFollow to mark OpenAI's stories across the site. No account needed.