Perplexity AI, the company behind the "Perplexity" answer engine, co-hosted the "U.S.–Japan AI Governance Symposium" together with Stanford University, the U.S.-Japan Foundation, and Kyoto University at SHIBUYA QWS in Tokyo on May 11, 2026[1]. Former Digital Minister and House of Representatives member Masaaki Taira delivered the keynote, and policymakers and industry leaders from both countries discussed how the challenges of AI governance have shifted from abstract principles to the implementation stage.
Leaders from Japan and the U.S. Gather in Shibuya
The symposium was organized by Stanford University and the U.S.-Japan Foundation, with Kyoto University and Perplexity AI serving as supporters and collaborators[1]. The venue was the SHIBUYA QWS Scramble Hall on the 15th floor of Shibuya Scramble Square, and it was held as an invitation-only, in-person event[1].
To encourage candid, high-level dialogue, the proceedings followed the Chatham House Rule, under which speakers are not identified[1]. Policymakers, the legal academic community, technology companies, and practitioners came together, and "Trust" in the entire process of how AI is governed and integrated into society emerged as the central theme running through every session[1].
Lawmaker Masaaki Taira's Keynote: Toward "the World's Most AI-Friendly Country"
The keynote was delivered by Masaaki Taira, a member of the House of Representatives who chairs the Liberal Democratic Party's project team on the evolution and social implementation of AI[1]. Centering on Japan's vision of becoming "the world's most AI-friendly country," Taira spoke about the importance of a uniquely Japanese governance model that balances flexible oversight that does not hinder technological innovation with the safety of society[1].
He also shared a broad outlook on the next steps that government, industry, and academia should pursue together to build a "trusted social foundation" that accounts for the new risks accompanying AI's evolution as well as international perspectives[1].
Three Key Points Raised in the Panels
The panel discussions distilled the conversation into three major points[1].
The first is regulatory convergence. While each country's approach differs by cultural background—Japan's flexible, agile model, Europe's legislative model, and the United States' market-driven model—the view was presented that approaches are now converging toward Japan's "flexible, trust-based model"[1]. The framing is that Japan occupies an important position as a "bridge" that mediates between the regulatory approaches of different countries[1].
The second is intellectual property. Against the uncertainty surrounding copyright, panelists argued that both the clarification of boundaries through litigation and practical cooperation through contracts—such as direct licensing and data-sharing infrastructure—are needed[1]. In particular, RAG (retrieval-augmented generation), which clearly indicates sources, drew expectations as a realistic solution that enables proper attribution and compensation[1].
The third is social implementation. Japan's AI adoption is at a turning point from PoC (proof of concept) to large-scale deployment, and in organizations that embed AI across their entire workflow, the human role is shifting toward oversight, judgment, and strategic decision-making[1]. To secure trust and privacy, a hybrid configuration combining local and cloud environments is considered crucial[1].
A Diverse Lineup of Speakers and the Conclusion of "Coordinated Action"
Each panel featured figures at the forefront of policy, law, and technology[1]. Panel 1 was moderated by Stanford University Professor Nate Persily and included a Cabinet Office AI policy planning official and House of Representatives member Akihisa Shiozaki[1]. Panel 2 was led by Kyoto University Project Professor Hiroki Habuka, where Waseda University Professor Tatsuhiro Ueno and a Rakuten Group executive discussed intellectual property and information integrity[1]. Panel 3, themed on U.S.–Japan cooperation and social implementation, was moderated by Perplexity AI Chief Business Officer (CBO) Dmitry Shevelenko, and included Sakana AI Co-Founder and CEO David Ha and Microsoft Japan's Chief Technology Officer[1].
The symposium concluded with the message that rather than relying on a single actor or regulatory approach, "coordinated action" in which government, AI providers, and the entire ecosystem move in step is the path to a trusted AI society[1].
Perplexity is an AI company founded in 2022 with the mission of "power the world's curiosity"[2]. Centered on its "Perplexity" answer engine, which responds with inline citations based on trusted information sources, it also offers "Comet Browser," an AI-native web browser, and "Perplexity Computer," a large-scale multi-model orchestration platform; the company is said to answer more than 1.5 billion questions worldwide every month[1][2].
Summary
The "U.S.–Japan AI Governance Symposium," co-hosted by Perplexity AI together with Stanford University and others, served as a venue showing that the AI governance debate is moving from abstract principles to practical implementation. The keyword running through the themes of regulatory convergence, coexistence with intellectual property, and the changing role of humans was "trust," and the importance of "coordinated action" linking government, industry, and research institutions was reaffirmed. Whether Japan can become the "bridge" that mediates among different countries' regulations looks set to be the focus going forward.
出典:https://prtimes.jp/main/html/rd/p/000000026.000157647.html
