OpenAI Offers Cyber AI Tool to UK Banks as Rival Anthropic Blocks Access

OpenAI has extended an offer to nine major UK banks to access its cyber security AI tool GPT-5.5 Cyber, a move that comes as its competitor Anthropic has blocked them from previewing its own version, Claude Mythos. This development highlights the growing importance of AI in cybersecurity, as both tools are designed to identify hidden security weaknesses in digital systems, often outperforming human capabilities in certain hacking and cyber-security tasks. The introduction of Mythos in April generated significant attention, particularly after Anthropic claimed it had uncovered a nearly 30-year-old vulnerability in a legacy system.

However, concerns have been raised by finance ministers, central bankers, and financiers regarding the potential for Mythos to compromise the security of financial systems. The AI Security Institute, which has evaluated both Anthropic's Mythos and OpenAI's GPT-5.5 Cyber, reported that they achieved a similar level of performance in tasks assigned to them. Bank of England Governor Andrew Bailey recently expressed concerns that UK banks were still unable to access Mythos to test their digital systems and apps.

Former UK Chancellor George Osborne, now a senior executive at OpenAI, stated that Bailey had not contacted him directly regarding access to GPT-5.5 Cyber. Osborne emphasized that OpenAI's goal is not to restrict access but to ensure these tools are in the right hands, adding, "The key things with these tools is that they need to be in the hands of the right people. We want to make sure that the forces that are establishing order in our democracies have these tools, and the forces that want to disrupt us or commit crime, do not." The UK banks that will gain access to OpenAI's GPT 5.5 Cyber include Lloyds Banking Group, HSBC, and Nationwide, while NatWest and Santander already have access through existing agreements.

According to Anthropic, the company is working urgently to expand access to Mythos, which it believes offers superior capabilities to 5.5 Cyber, necessitating more caution. Initially, Anthropic provided Mythos to a collective of 42 companies, primarily other US tech firms. In contrast, OpenAI has been more expansive with access to Cyber 5.5, opening it up in the EU, as well as to banks in Japan and Canada.

Both firms operate on a usage-based pricing model, although Anthropic allocated $100 million towards Mythos previews. Professor Alan Woodward, a cybersecurity expert from Surrey University, noted, "I don't necessarily expect these tools to surface things humans wouldn't find eventually but they are relentless and incredibly thorough in sorting through the millions of lines of code which are in banking apps alone." He added that while AI can process vast amounts of data quickly, human oversight is still necessary to verify findings and address false positives. The development underscores the critical role AI is set to play in enhancing cybersecurity, particularly in the financial sector, where legacy systems and the vast scale of digital infrastructure present significant challenges.