Large Japanese banks reported a twofold increase in identified system vulnerabilities as of August compared to the previous year. This trend follows the introduction of the Claude Mythos AI model by Anthropic in April, which allows for faster detection of security flaws.
Financial institutions in Japan gained access to the Mythos model this summer. However, the situation became complex due to US government requirements to suspend access for foreign users. Banks are currently evaluating the tool to identify and patch security gaps.
In May, the Japan Financial Services Agency (FSA) advised financial groups to strengthen their defense measures. The regulator suggested that market participants consider voluntarily suspending critical services to prevent the spread of damage from potential cyberattacks.
The Financials ISAC Japan organization is currently developing protection guidelines for banks. This initiative is being conducted at the request of the FSA and the Bank of Japan to provide a framework for responding to cyber threats.
