“Allied on AI Assurance: Technology & Policy to Enhance AI Security”

From December 8-13, US-Japan NEXT Alliance Initiative Senior Director Jim Schoff led a six-person US delegation of Artificial Intelligence (AI) experts from the national research community and policy experts to meet with counterparts in Tokyo. The purpose of the trip was to foster networking, share information, and explore opportunities for bilateral research collaboration to strengthen AI defenses, bolster cybersecurity, improve data quality, and help mitigate disinformation.

The delegation on this AI Security study trip met with policy makers and AI researchers in Japan’s government, private sector, and research institutes to discuss strategic approaches and develop joint initiatives for AI security. The US experts came from Lawrence Livermore National Laboratory, MITRE, the Nuclear Threat Initiative, and RAND, meeting with Japanese officials and researchers from the Cabinet Office, various ministries, Japan’s AI Safety Institute, and Japanese universities and laboratories. The trip culminated with a large public event featuring the US delegation at the University of Tokyo, Research Center for Advanced Science and Technology (RCAST).

Background:

A phenomenon some call “adversarial AI” has become an economic and national security concern not only in terms how it can threaten cybersecurity efforts or AI training model integrity, but also its potential to spread disinformation and ultimately dampen AI utilization.

As the 2025 US AI Action Plan notes, the inner workings of frontier AI systems are poorly understood, lack predictability, and are susceptible to adversarial inputs in ways that can undermine national security. At the same time, nations are trying to avoid regulating AI development in ways that inhibit innovation. The US and Japan are wrestling with these challenges individually and collectively through multilateral forums such as the G7 and OECD, but bilateral networking on this topic can be an important supplement, especially for the technology behind policy development.

Bilateral collaboration in AI fields is emerging through such channels as the Cross-Pacific AI Initiative (X-PAI) involving US and Japanese universities and companies, and Japan’s National Institute of Information and Communications Technology’s (NICT) CREATE program supporting cooperation with US firms and universities. Our trip created opportunities to share best practices, compare and align strategies, and explore cooperative projects or investments, as Japan worked to finalize its AI Basic Plan by the end of 2025.

• For an outline of the NEXT Alliance AI Study Trip and delegation members, click here.
• To read the summary of the study trip authored by NEXT Alliance Initiative Senior Director Jim Schoff, click here.

Day One:

The US delegation carried out a series of meetings with AI and economic security-related officials at Japan’s National Security Secretariat to discuss Japan’s AI priorities and programs, followed by a meeting with US Embassy officials to learn about US priorities for AI cooperation with Japan. Later in the afternoon, Sasakawa USA-hosted the roundtable US-Japan Expert Dialogue on Physical AI Security with Japanese expert scholars and private sector experts to explore policy and technology issues involved in the expanding nexus of virtual and physical systems leveraging AI.

The first day was capped off by a welcome reception hosted by US Embassy Deputy Chief of Mission Aaron Snipes, which allowed delegates to build creatively on the foundation of the day’s initial engagements. The reception provided the US delegation with an opportunity to meet US and Japanese AI-interested scholars and officials in a casual setting for networking.

Day Two: 

Sasakawa USA co-organized with the National Institute of Information and Communications Technology (NICT) a roundtable US-Japan Expert Dialogue on Mitigating Cyber Misuse Risks for AI Systems with AI researchers and private sector representatives sharing their perspectives about AI safety, cybersecurity, and identifying promising opportunities for bilateral collaboration. The US delegation also made a site visit to Japan’s AI Safety Institute to exchange expertise and learn about Japan’s role in convening AI Safety Institutes around the world through the Hiroshima AI Process, explored the Ueno area, and concluded the day with a sake tasting/lecture dinner alongside several meeting partners.

Day Three:

Sasakawa USA arranged for three members of the US delegation to participate in a public event on US-Japan Perspectives on Risks from Frontier AI hosted by the University of Tokyo’s Research Center for Advanced Science and Technology (RCAST) and its Economic Security Intelligence Lab (ESIL), attended in person by over 100 people. Later in the day the US delegation conducted a policy-focused AI safety roundtable with the AI Office in the Secretariat of Science, Technology, and Innovation Policy at the Cabinet Affairs Office to discuss Japan’s approach to AI safety and security within the 2025 AI Basic Plan.

Concluding Dinner: 

Before the US delegation returned home, Sasakawa USA hosted a farewell dinner involving some of the friends we had made during the trip and other leaders in Japan’s science diplomacy and technology research communities.

Photos: