OpenAI offers UK banks access to GPT-5.5 Cyber amid Anthropic's blocked Mythos

OpenAI has granted nine major UK banks access to its cyber security AI tool GPT-5.5 Cyber, while blocking them from using Anthropic's Claude Mythos. Mythos sparked concern when Anthropic claimed it detected a 30-year-old vulnerability in a legacy financial system, prompting fears it could compromise financial infrastructure. The AI Security Institute found both GPT-5.5 Cyber and Mythos achieved similar performance in identifying hidden security flaws. Bank of England governor Andrew Bailey confirmed UK banks still lack access to Mythos for testing their digital systems. This divergence highlights growing tensions between AI security tools and potential risks to financial stability.