Japan’s three major banking groups, including MUFG Bank, are expected to gain access to U.S. AI startup Anthropic’s latest artificial intelligence model, Claude Mythos, sources familiar with the matter said on Wednesday, News.Az reports, citing Kyodo.

The development follows reports that the Japanese government sought access to the AI model as Japan and the United States strengthen cooperation to respond more rapidly to cyberattacks targeting vulnerabilities in financial systems.

Claude Mythos, released in April by Anthropic, has been widely seen as a major challenger in the AI sector long dominated by OpenAI and its ChatGPT platform. However, the model has also raised concerns that it could potentially be exploited to attack various digital systems.

Executives from MUFG Bank, Sumitomo Mitsui Banking Corporation and Mizuho Bank attended a meeting with Scott Bessent during his visit to Japan earlier this week, with access to Claude Mythos believed to have been among the issues discussed.

Due to safety and security concerns, access to the Claude Mythos model is currently limited to selected IT companies, including Google, as well as certain financial institutions. Anthropic has stated that the preview version of the model “has already found thousands of high-severity vulnerabilities.”

Japan’s Financial Services Agency has decided to establish a framework bringing together public- and private-sector officials to discuss stronger responses related to Claude Mythos, in coordination with the Bank of Japan and major commercial banks.

According to the sources, a working group could begin meeting as early as Thursday to discuss information-gathering efforts and measures to strengthen cybersecurity defenses.

Anthropic was founded in 2021 by former OpenAI employees.

News.Az 

By Nijat Babayev