UK banks blocked from cyber AI tool Mythos get offer from rival OpenAI (www.bbc.com)

OpenAI has stepped in to offer its cybersecurity AI tool, GPT-5.5 Cyber, to nine major UK banks after rival Anthropic blocked them from accessing its highly anticipated tool, Claude Mythos. Both AI systems are engineered to unearth hidden vulnerabilities in digital infrastructures, with reports indicating they perform similarly in tasks related to cybersecurity. Mythos gained attention earlier this year for revealing a vulnerability in a legacy system that had been undetected for nearly three decades. Meanwhile, Bank of England Governor Andrew Bailey expressed concern over UK banks' limited access to advanced security testing tools, highlighting the urgency of equipping financial institutions with these technologies. OpenAI's former CEO, George Osborne, emphasized that the firm aims to provide Cyber 5.5 to organizations that would use it responsibly, ensuring that malicious actors do not gain access. The banks benefiting from this offer include major players like Lloyds, HSBC, and Nationwide, while NatWest and Santander already possess access through existing agreements. Anthropic is now working to expand access to Mythos but maintains a cautious approach, believing its tool surpasses Cyber 5.5 in capability. Experts indicate that while these AI tools can dramatically reduce the time needed to identify coding issues in banking applications, human oversight remains essential due to the potential for false positives in their findings.