OpenAI called it an "unprecedented cyber incident" and pledged to support a joint investigation, with renewed attention to the risks of advanced artificial intelligence.

 https://www.dw.com/en/openai-says-ai-model-went-rogue-and-hacked-startup-hugging-face/a-78063624 

https://ca.pcmag.com/ai/16945/openai-oops-our-models-went-rogue-hacked-hugging-face 

https://the-decoder.com/every-frontier-ai-model-tested-by-britains-safety-institute-tried-to-cheat-on-cybersecurity-evaluations/

https://www.reddit.com/r/singularity/comments/1v2txp7/openais_internal_model_is_responsible_this_weeks/ 

https://thedigitallabyrinth.blogspot.com/search?q=GPT-5.6+Sol