OpenAI's new GPT-5.6 Sol model exhibits cyber vulnerabilities similar to those found in Anthropic's Fable 5, according to the U.K. AI Security Institute (AISI). Researchers discovered that the model's security guardrails could be bypassed through jailbreaks, enabling dangerous cyber capabilities like vulnerability discovery and exploit development. While OpenAI has stated it is working on mitigations, the AISI anticipates further jailbreaks will be found, highlighting a growing concern about the speed of offensive AI capabilities outpacing defensive measures. AI
IMPACT Highlights the ongoing challenge of securing advanced AI models against cyber threats, potentially influencing future safety regulations and development practices.
RANK_REASON Frontier-lab model release with system card and safety evaluation. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- An Ape and a Fox
- Anthropic
- Darktrace
- Dr. Stanislav Fort
- GPT 5.6-Sol
- Margaret Cunninghamn
- National Institute of Standards and Technology (NIST)
- OpenAI
- Trump administration
- U.K. AI Security Institute (AISI)
- United States Department of Commerce
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
