OpenAI has recently extended access to its advanced GPT-5.5 model to select financial institutions in Japan, a move aimed at bolstering defenses against cyber threats. This development was announced by Satsuki Katayama, Japan's Finance Minister, following her discussions with Jason Kwon, OpenAI's chief strategy officer, in Tokyo on Friday.
The advent of sophisticated AI models capable of high-level coding has significantly enhanced the capabilities of cybercriminals, enabling them to pinpoint and exploit cybersecurity vulnerabilities more effectively. Reacting to this escalating risk, banks and other organizations are now banking on early access to these cutting-edge technologies, which are reserved for a select group of trusted partners.
Although specific financial institutions were not disclosed, Katayama highlighted that the access represents a significant advancement in fortifying Japan's financial sector against potential cyberattacks. According to a report from the Nikkei newspaper, major banks in Japan, including MUFG Bank, Sumitomo Mitsui Banking Corp, and Mizuho Bank, are anticipated to be among those receiving this crucial access, positioning them competitively alongside rival firms, such as Anthropic's Claude Mythos.
While MUFG and Mizuho have opted not to comment on the matter, Sumitomo Mitsui Banking Corp was not immediately available for information. Katayama mentioned that the Japanese government and financial entities are also poised to utilize Mythos to further enhance their cybersecurity measures.
In a proactive response to AI-related threats, Japan has recently formed a collaborative public-private working group this month, dedicated to tackling cybersecurity challenges within the financial sector connected to advancements like Mythos.
The decision to grant access to these AI systems was the result of negotiations between the Japanese and U.S. governments, following similar arrangements made for European firms.
