GPT-5.6 safety tests expose jailbreak cracks. This localization keeps the verified facts and reads the story through the lens of product, infrastructure and governance.
What happened
GPT-5.6 safety tests expose jailbreak cracks is the current news peg. The verified record centers on these points:
- UK AISI received pre-deployment access to GPT-5.6 Sol for expert cyber tasks
- Sol scored 95.0% ± 9.8% on AISI CTF expert tasks versus 85.0% ± 11.6% for GPT-5.5
- in The Last Ones 32-step range, Sol completed 7 of 10 runs versus 2 of 10 for GPT-5.5
- OpenAI says GPT-5.6 Sol safeguards blocked about 10x more potentially harmful cyber activity
Why it matters
The broader signal is agentic cyber risk is now about multi-step task completion after inducement, not only one-turn refusal.
What to watch
The next check is how model providers combine identity, environment authorization, audit trails and post-release red teaming.
Sources：Fortune, OpenAI GPT-5.6 System Card, Anthropic notes, CocoLoop.