Security & Privacy13 min
GPT-5.5 Cyber Eval: AISI Finds Parity with Mythos 2026
UK AI Security Institute's April 30 GPT-5.5 cyber eval reveals parity with Claude Mythos on expert CTF tasks and the 32-step Last Ones attack range.
Tag
2 posts
Security & Privacy13 min
UK AI Security Institute's April 30 GPT-5.5 cyber eval reveals parity with Claude Mythos on expert CTF tasks and the 32-step Last Ones attack range.
Security & Privacy11 min
The UK AI Security Institute's Claude Mythos evaluation: 73% on expert CTFs, first model to autonomously complete a 32-step enterprise network attack.