How Japan’s new AI Act fosters an innovation-first ecosystem

In May 2025, Japan enacted a landmark piece of legislation — the Act on the Promotion of Research, Development and Utilisation of Artificial Intelligence-Related Technologies — with a clear ambition: to make AI the foundation of Japan’s economic revival and digital leadership. This law does more than set policy direction; it marks a philosophical departure from the dominant regulatory approaches shaping the global AI landscape. At a time when major regions like the European Union are moving toward risk-based regimes, Japan’s AI law signals a pivot toward coordination, and voluntary responsibility.

The Japanese approach

The contrast could not be starker. The European Union’s AI Act, passed in 2024, is defined by its restrictive architecture. It classifies AI systems into risk tiers — ranging from “Unacceptable” to “Minimal” — and imposes strict, legally binding obligations on developers, especially those building high-risk applications in areas such as health, education, employment, or law enforcement. The EU framework is comprehensive, enforceable, and aligned with its values of human dignity and digital sovereignty. Non-compliance invites steep penalties and regulatory scrutiny.

Japan’s approach is fundamentally different. Rather than focusing on risk classification and penalties, it emphasises enabling innovation, encouraging collaboration, and fostering international competitiveness. The law creates an Artificial Intelligence Strategy Headquarters under the Cabinet and tasks it with formulating and implementing a national Basic Plan for AI. This plan will cover everything from foundational research to industrial deployment, international cooperation and public education.

Crucially, the Japanese law avoids the trap of overregulation. It does not create binding enforcement mechanisms or define risk categories. Instead, it frames AI-related technologies as foundational for societal development, economic growth, administrative efficiency, and national security. The state assumes responsibility for facilitating research, creating shared infrastructure, supporting workforce development, and ensuring transparency and ethical conduct in AI utilisation. Local governments, universities, research bodies, businesses, and even the public are assigned cooperative roles under the law’s basic principles.

This model rests on two key assumptions. First, that innovation ecosystems thrive better in the absence of rigid regulatory burdens. Second, that voluntary cooperation — when guided by national coordination and ethical principles — can effectively mitigate risks associated with the misuse of AI. Article 13 of the Act affirms the government’s responsibility to develop guidelines that reflect international norms and prevent harm, such as misuse, privacy breaches, or intellectual property violations. However, it stops short of codifying hard rules or penalties.

The strengths of this approach are obvious. Japan avoids the chilling effect that often accompanies over-regulation. It builds an innovation-first ecosystem, where AI development can progress across sectors — public and private — without being prematurely constrained by legal ambiguity or bureaucratic friction. It also signals to industry and academia that the government is a facilitator, not a regulator.

Myriad challenges

But there are risks too. In the absence of clear standards and enforcement, critical questions remain: what happens when AI harms go unreported? How do we define accountability in the event of bias, disinformation, or algorithmic failures? How will Japan ensure that voluntary principles translate into enforceable safeguards in sectors such as healthcare or defence?

By avoiding a risk-tiered model like that of the EU, Japan may gain agility — but at the potential cost of clarity and public trust. As generative AI and autonomous systems become more embedded in daily life, even jurisdictions that adopt light-touch approaches will eventually face mounting pressure to articulate what “responsible AI” means not just in theory, but in law.

The geopolitical context also matters. The EU’s model is shaped by its strong tradition of rights-based governance and a cautionary approach to data and digital technologies. Its AI Act is a natural extension of its General Data Protection Regulation-era regulatory posture. Japan, on the other hand, is facing unique economic challenges — a shrinking workforce, global competition in advanced technologies, and the need to stimulate domestic innovation. The new AI law reflects a strategic choice to double down on science and technology as national growth drivers.

That does not mean Japan is ignoring international alignment. On the contrary, Article 17 of the law mandates that the state actively engage in international cooperation and norm-setting. This is a timely move. Just as the Financial Stability Board (FSB) is conducting a global peer review on crypto frameworks, similar coordination efforts are emerging in the AI space — under the G7 Hiroshima Process, Organisation for Economic Co-operation and Development frameworks, and the UN’s AI advisory body. Japan’s ambition to lead in these forums will require it to balance its promotion-first model with a willingness to define guardrails in line with emerging global standards.

Global methods

Other countries are also taking diverse approaches. In the U.S., the momentum is shifting toward legislative clarity through the AI Disclosure Act. This aims to delineate agency jurisdiction, ensure transparency in training data and outputs, and safeguard national security interests. The U.S. approach, while still evolving, seeks a balance between innovation and oversight — empowering sectoral agencies to issue context-specific rules for AI deployment.

Meanwhile, the United Arab Emirates (UAE) is positioning itself as a leader in state-led AI strategy. With its Office of Artificial Intelligence, national AI university, and industry-driven AI sandbox programmes, the UAE blends strategic investment with targeted regulation. Sectoral pilots in education, transport, and healthcare have helped create trusted ecosystems while still fostering AI-led transformation. Unlike Japan’s voluntary model, the UAE’s approach is executive-driven but agile and business-friendly.

In many ways, Japan’s AI law is a gamble on institutional trust. It bets that government ministries, research institutions, local authorities, and businesses can work together to ensure ethical AI innovation — without needing to be policed into compliance. This reflects a broader cultural confidence in technocratic leadership and consensus-driven governance. But this trust must be earned continuously. The law’s promise will only be fulfilled if the Artificial Intelligence Strategy Headquarters can effectively coordinate across sectors, issue timely guidance, and revise its policies based on real-world feedback and global developments. The law itself includes provisions for future review and amendment — a tacit acknowledgement that today’s principles may need tomorrow’s precision.

The world is watching to see whether Japan’s model of responsibility without rigidity can truly offer a sustainable and scalable path forward. If it succeeds, it could offer a compelling alternative to both laissez-faire deregulation and enforcement-heavy regimes. But if it falters, it will serve as a cautionary tale about the risks of moving too lightly in the face of transformative technologies. Japan has chosen to lead with coordination, not control. The real test begins now.

Sanhita Chauriha is a Technology Lawyer.