MUFG Bank, Sumitomo Mitsui Banking Corp. and Mizuho Bank are set to gain access to Anthropic's AI model Mythos, according to a person familiar with the matter.
The move follows Mythos's limited release last month, which triggered concern that the model could usher in a new wave of cybersecurity risks.
Anthropic developed Mythos in the US, and the planned access for Japan's three biggest banks was earlier reported by Nikkei.
A new AI can hack almost anything. Is giving it to banks a real solution, or just the start of a new arms race?
Anthropic's AI is deemed too dangerous for public release but scores perfectly on safety tests. How secure can our digital world truly be?
Is the world's most powerful hacker AI a true cyber-weapon, or a brilliant marketing ploy to sell advanced security solutions?