Japan's government and several domestic banks have been granted access to Anthropic's latest Claude Mythos artificial intelligence model, which is equipped with advanced cybersecurity capabilities, according to an official government announcement on Wednesday [1]. This move comes amid growing concerns about cyberattacks driven by sophisticated AI technologies [1]. Finance Minister Satsuki Katayama emphasized the importance of this development, stating, 'There is no doubt that this will contribute to strengthening security' and highlighting that it is 'unacceptable' for Japanese financial institutions to lag behind other countries in cybersecurity measures [1].
Hisashi Matsumoto, the minister in charge of cybersecurity, also praised the decision, describing it as a 'significant step forward' and pledging to take further actions to 'strengthen the cyber resilience of Japan as a whole' [1]. The expansion of access rights by Anthropic to include the Japanese government and financial institutions was noted as a key milestone, with Matsumoto referencing the opening of the 'door' to like-minded countries [1].
The development follows Katayama's statement last week that major Japanese banks had also been granted access to OpenAI's newest GPT-5.5-Cyber model, which is considered to have similar advanced cybersecurity features as Claude Mythos [1]. Leading banks such as MUFG Bank are expected to utilize the Anthropic model, according to a source close to the matter [1].
The granting of access to Claude Mythos was anticipated after U.S. Treasury Secretary Scott Bessent indicated during his visit to Japan through May 13 that it would occur within two weeks, as confirmed by Katayama [1]. Initially, Anthropic had restricted access to the model, which is highly capable of identifying security vulnerabilities, to IT companies like Google LLC and select financial institutions after its unveiling in April, due to concerns about potential misuse for cyberattacks [1].
CONCLUSION
The Japanese government and major banks' access to Anthropic's Claude Mythos AI model marks a significant advancement in the nation's cybersecurity efforts. This move is expected to enhance the cyber resilience of Japan's financial sector and aligns with global trends in adopting advanced AI for security purposes.