Japan’s three largest banks are slated to gain access to Mythos, the artificial intelligence model produced by Anthropic, in roughly two weeks, according to a person with direct knowledge of the arrangement.
Cybersecurity specialists have raised significant concerns about Mythos and its interaction with the banking sector and existing legacy technology stacks. Those concerns have prompted warnings from regulators and policymakers, reflecting unease about operational and security implications as the model is adopted in financial services.
To date, a number of U.S. banks have been permitted to use Mythos, and Anthropic has indicated plans to broaden access to banks in Europe and the U.K., among other organisations. The expected rollout to Japan’s largest banks is part of that wider expansion.
Mitsubishi UFJ Financial Group, Mizuho Financial Group and Sumitomo Mitsui Financial Group were each contacted but declined to comment on the expected access. Anthropic did not respond to a request for comment.
On a related policy track, Japan’s Finance Minister Satsuki Katayama met with U.S. Treasury Secretary Scott Bessent on Tuesday. After that meeting, Katayama said Japan would establish a public-private working group this week to examine cybersecurity risks to the Japanese financial system posed by Mythos. The working group is scheduled to convene for the first time on Thursday.
The development underscores a tension between expanding access to advanced generative AI models and safeguarding financial infrastructure. Market indicators for major Japanese banking tickers referenced in public market displays included 8306, 8316 and 8411.
Context and implications
- The imminent access for Japan’s three largest banks represents another step in Anthropic’s broader move to supply Mythos to financial institutions outside the United States.
- Regulatory attention has intensified as officials and cybersecurity experts flag risks tied to integrating the model with legacy systems, prompting coordinated public-private responses.