Japan’s government and some of the country’s largest financial institutions have secured access to Claude Mythos, Anthropic’s highly advanced cybersecurity-focused artificial intelligence model, as regulators and banks intensify efforts to strengthen cyber resilience against increasingly sophisticated digital threats.
The move places Japan among a select group of countries granted access to the frontier AI system, which has attracted global attention for its ability to identify previously undiscovered software vulnerabilities. Anthropic had previously restricted access to the model, citing concerns that its capabilities could be misused if widely released.
According to reports, Japan’s three largest banking groups, MUFG Bank, Mizuho Bank and Sumitomo Mitsui Banking Corporation, are among the financial institutions receiving access to the technology. The announcement coincides with Anthropic’s decision to expand the Mythos programme to approximately 150 organisations across more than 15 countries.
Cybersecurity Becomes a Strategic Priority
The deployment comes as financial institutions face mounting cyber risks from increasingly capable threat actors and AI-powered attack techniques. By leveraging Claude Mythos, banks are expected to identify vulnerabilities across critical systems faster and remediate security weaknesses before they can be exploited.
Japan’s Finance Minister and Minister of State for Financial Services, Satsuki Katayama, highlighted the importance of strengthening cybersecurity across the financial sector, noting that financial institutions remain among the most frequent targets of economic crime and cyberattacks.
Katayama said maintaining strong cyber defences is critical not only for individual institutions but also for preserving confidence in Japan’s broader financial system. She also emphasised that cybersecurity standards within Japanese financial services must remain competitive with those of leading U.S. financial institutions as the sector continues to play an important role in the country’s economic growth strategy.
Regulators Prepare for the Frontier AI Era
Japan’s Financial Services Agency (FSA) has already begun preparing financial institutions for the implications of frontier AI technologies. Last month, the regulator urged banks and financial firms to implement immediate cybersecurity measures under the direct oversight of senior management.
The agency warned that advanced AI systems could uncover large volumes of previously unknown vulnerabilities, creating operational challenges for institutions that may struggle to address every issue simultaneously. As a result, the FSA encouraged firms to prioritise remediation efforts and allocate resources strategically rather than attempting to expand IT teams in the short term.
The regulatory guidance reflects a growing recognition that AI is becoming both a cybersecurity defence tool and a source of new operational challenges for financial institutions.
OpenAI Also Expands Cybersecurity Access
Japan’s banking sector is also attracting attention from other leading AI providers. During a recent meeting with Katayama, OpenAI Chief Strategy Officer Jason Kwon reportedly confirmed that selected Japanese financial institutions would receive access to GPT-5.5 Cyber, OpenAI‘s latest cybersecurity-focused AI model.
The parallel rollout of advanced cybersecurity AI from both Anthropic and OpenAI signals growing competition among AI vendors to become trusted partners for governments, regulators and critical infrastructure operators.
For banks, access to these specialised systems could represent a significant shift in how cybersecurity programmes are managed, enabling organisations to identify and address vulnerabilities at a scale and speed that traditional security teams have struggled to achieve.
What this means for the industry
- Banks are becoming early adopters of frontier AI designed specifically for cybersecurity rather than customer-facing applications.
- Regulators are increasingly encouraging financial institutions to use AI as a defensive tool against evolving cyber threats.
- AI vendors are competing to establish themselves as strategic cybersecurity partners for governments and financial institutions.
- The ability to discover unknown vulnerabilities may become a major differentiator in future bank cybersecurity programmes.
- Cyber resilience is emerging as one of the most important use cases for advanced AI within financial services.
