Japanese megabanks to access Anthropic's Mythos AI amid cybersecurity concerns
Japanese megabanks Mitsubishi UFJ, Mizuho Financial Group and Sumitomo Mitsui were set to deploy Anthropic's Mythos AI model, which can identify vulnerabilities in enterprise systems. Cybersecurity experts and regulators raised concerns about potential misuse if the technology were compromised, prompting Japan's Finance Ministry to establish a working group on the risks.
What happened
Japan's three largest banks—Mitsubishi UFJ Financial Group, Mizuho Financial Group and Sumitomo Mitsui Financial Group—were reportedly set to gain access to Anthropic's Mythos AI model in May 2026.
During testing, Mythos uncovered vulnerabilities across operating systems, browsers and enterprise software. The AI's ability to aggressively identify weaknesses in older enterprise systems raised concerns from regulators and cybersecurity experts about potential misuse if the technology fell into the wrong hands.
Reported impact
Cybersecurity experts viewed Mythos as posing significant challenges to the banking industry and its legacy technology systems.
Japan's Finance Minister Satsuki Katayama announced the country would launch a new public-private working group focused specifically on cybersecurity risks tied to the model.
Index analysis
Banks sought access to Mythos for its capability in handling complex cybersecurity analysis, automation and infrastructure management as cyberattacks become more sophisticated.
Governance controls
Japan's Finance Ministry established a public-private working group to address cybersecurity risks associated with Mythos deployment.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.