(Excerpt)
(Tuesday, May 12, 2026, ９:57 am to 10:10 am)
[Opening remarks:]
- Minister)
-
Yesterday, I had dinner for about two hours with U.S. Treasury Secretary Bessent, who is visiting Japan. Today, I also held a meeting with him for 35 minutes, from 9:20 a.m. to 9:55 a.m.
At the meeting,... with regard to responding to cyber threats associated with advances in AI, I first stated that it is important for allies to respond in concert, and he concurred. We also discussed coordinating broadly at the international level, including strengthening the resilience of supply chains for critical minerals, in view of the G7 meeting taking place this week.
On that basis, we reaffirmed that Japan and the United States would continue to work closely together and coordinate with each other in areas where cooperation is possible, with a view to addressing these issues, including bilaterally and at forums such as the G7 meeting in Paris starting next Sunday. I will attend the G7 meeting with Secretary Bessent, and it will be our next immediate opportunity as finance ministers. This was my fifth face-to-face meeting with Secretary Bessent, and we have spoken in a variety of settings. Considering that they will be visiting China, I believe it was extremely meaningful, and indeed very special, that we were able to spend a considerable amount of time during Secretary Bessent’s visit to Japan having in-depth discussions.
In addition, we will launch a working group under the Public–Private Coordination Meeting on Strengthening Cybersecurity Measures in the Financial Sector Against AI-Related Threats. At the Public–Private Coordination Meeting on Strengthening Cybersecurity Measures in the Financial Sector Against AI-Related Threats held on April 24, I proposed the establishment of this working group and received the support of the participants. We are scheduled to launch it on May 14.
I expect this working group to deepen discussions at the working level so that the financial industry, IT companies, the government, the Bank of Japan, and other relevant parties can develop a common understanding of the threats posed by advances in AI technologies and consider possible responses. The United States has also been very cooperative on this point. It appears that the U.S. government, through its work to date, has developed a certain degree of understanding and cooperation with entities that possess such highly advanced AI, primarily Anthropic, and also OpenAI and Google, among others. We are moving in the direction of having the U.S. side share that information with us and working together.
[Questions and answers:]
- Q.
-
(In your meeting with Secretary Bessent) Did you discuss the impact of Anthropic’s Mythos AI model on the financial system and possible responses?
- A.
-
There is already recognition that AI technologies associated with roughly three major players — Anthropic, with its Mythos, OpenAI, with what I believe is said to be its forthcoming GPT-6, and Google — are extremely advanced and can pose certain risks, given their ability to identify vulnerabilities in programs almost instantly. I was told that, over the past two weeks or so, considerable progress has been made in developing protocols between these entities and the U.S. Treasury. I think it is very positive that such matters are understood in that way in their home country. Frankly, once such capabilities emerge, China could catch up within six months or a year. However, that has not yet happened. I believe we shared the view that Western partners must coordinate closely and take steps to prevent such technologies from being weaponized by non-like-minded parties.
