By Leika Kihara
TOKYO, Could 29 (Reuters) – OpenAI has given some Japanese monetary establishments entry to its GPT-5.5 mannequin to assist forestall cyberattacks, Japanese Finance Minister Satsuki Katayama mentioned on Friday after a gathering with the U.S. firm’s chief technique officer.
The discharge of recent fashions of AI in a position to code at a excessive degree has given hackers an unprecedented potential to establish cybersecurity dangers and devise methods to use them.
One defence for the banks and different corporations susceptible to being destablilised is early entry to the brand new fashions which are out there solely to trusted companions.
With out naming any of the monetary establishments, Katayama mentioned the entry they have been getting was “an enormous step ahead in strengthening Japanese monetary establishments’ potential to defend in opposition to cyberattacks”. She was chatting with reporters after assembly Jason Kwon, OpenAI’s chief technique officer, in Tokyo.
The Nikkei newspaper reported on Thursday that Japan’s three greatest banks – MUFG Financial institution, Sumitomo Mitsui Banking Corp and Mizuho Financial institution – have been anticipated to achieve entry to OpenAI’s newest mannequin, which is believed to be on a par with that of rival Anthropic’s Claude Mythos.
MUFG and Mizuho declined to remark. Sumitomo Mitsui Banking was not instantly out there to remark.
Katayama mentioned on Friday that Japan’s authorities and monetary establishments have been additionally anticipated to achieve entry to Mythos to reinforce their defences.
Because it strives to counter threats from AI, Japan this month arrange a public-private working group to handle cybersecurity dangers to the monetary system posed by Mythos.
OpenAI’s choice to grant entry to Japanese monetary establishments adopted negotiations between Japanese and U.S. governments after it allowed European corporations comparable entry.
(Reporting by Leika Kihara; Enhancing by Himani Sarkar and Barbara Lewis)
