It would be interesting to read the post mortem (if they will release). In particular in which phase of the training it escaped the sandbox, in which way it used the information. Training is different from inference, I doubt there is a context there and where would the model store these information? In the weights?! That makes little sense to me
OpenAI cyber models broke out of training environment to hack Hugging Face
https://www.cnbc.com/2026/07/22/open-ai-cyber-models-hack-hugging-face.html
7 Comments
MagicShel@lemmy.zip · 7 pts · 11d
If you train a model for cyber warfare, you can't be surprised when it attacks the defenses keeping it from engaging in cyber warfare — even if they are your own.
lavander@lemmy.dbzer0.com · 0 pts · 11d
Oooorrrr there is more to the story
MagicShel@lemmy.zip · 1 pts · 11d
leaky_shower_thought@feddit.nl · 2 pts · 11d
the incident wasn't the "cheating" part. llms are known to scour the web to look for possible solutions. i think the incident was it found a solution in another website and finding that was made by the model itself.
meanwhile, no mention of the prompt guy or the prompt.
goatinspace@feddit.org · 2 pts · 11d
dparticiple@sh.itjust.works · 2 pts · 11d
Simon Willison's blog has some useful insights into what might have happened here: https://simonwillison.net/2026/Jul/22/openai-cyberattack/
lavander@lemmy.dbzer0.com · 1 pts · 11d
Ok this would be at inference time (solving the challenges) as opposed to the training that was described in the article.
This makes things more believable