OpenAI has provided nine UK banks with access to its GPT-5.5-Cyber model, filling a gap left by Anthropic's decision to restrict its Claude Mythos Preview from British and European financial institutions. Released on May 7, 2026, GPT-5.5-Cyber is part of OpenAI's Trusted Access for Cyber program, designed for pre-approved European financial firms. The model offers capabilities such as simulating attack vectors and analyzing live malware samples. The UK AI Security Institute tested GPT-5.5 in simulated cyberattacks, where it succeeded in 2 out of 10 trials, compared to Anthropic's Mythos, which achieved a 30% success rate. Despite this, UK banks remain unable to access Mythos, a situation highlighted by Bank of England Governor Andrew Bailey. OpenAI's move targets nearly all of the UK's systemically important banks, emphasizing the growing importance of AI-powered cybersecurity in the financial sector.