Las pruebas de seguridad de GPT-5.6 revelan grietas de jailbreak. Esta versión conserva los hechos verificados y los ordena desde producto, infraestructura y gobernanza.
Qué ocurrió
Las pruebas de seguridad de GPT-5.6 revelan grietas de jailbreak 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
Por qué importa
La señal de fondo es agentic cyber risk is now about multi-step task completion after inducement, not only one-turn refusal.
Qué observar
El próximo punto a comprobar es how model providers combine identity, environment authorization, audit trails and post-release red teaming.
Fuentes：Fortune, OpenAI GPT-5.6 System Card, Anthropic notes, CocoLoop.