OpenAI on X: "We recognize there are a lot of questions and speculative details circulating related to the Hugging Face incident. This is an unprecedented incident, and we think it marks an important moment for AI safety. We are still conducting a thorough review along with external advisors and with oversight from our Safety and Security Committee. Once the review is complete, we plan to publish a technical report of our learnings in the coming weeks." / X

Post

OpenAI
@OpenAI
Jul 21
We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation. Sharing preliminary findings to help defenders understand emerging risks:
OpenAI and Hugging Face partner to address security incident during model evaluation
From openai.com
OpenAI
@OpenAI
We recognize there are a lot of questions and speculative details circulating related to the Hugging Face incident. This is an unprecedented incident, and we think it marks an important moment for AI safety. We are still conducting a thorough review along with external advisors and with oversight from our Safety and Security Committee. Once the review is complete, we plan to publish a technical report of our learnings in the coming weeks.
12:41 AM · Jul 25, 2026363.1KViews
ram.
@ram_dawgg
Jul 25
Next time I obliterate my wife’s left overs at midnight and she gets mad ima hit her with the “this is an unprecedented incident, and we think it marks an important moment for marriage safety”
Harlan Stewart
@HumanHarlan
Jul 25
Unprecedented but not a surprise. Sam Altman himself predicted all the way back in 2015 that AI containment measures would stop working as AI became more powerful
Harlan Stewart
@HumanHarlan
Jul 23
“I’m not very optimistic that any of this will work” That’s what Sam Altman said in 2015 about using strong security measures such as airgapped computers to contain a superintelligent AI.
Riley Coyote
@RileyRalmuto
Jul 25
the unprecedented part is just that it happened publicly. we both know much, much, *much* wilder events have occurred with your models. internally. c'mon now.