When it comes to regulating artificial intelligence, Japan is often described as taking a “soft law” approach. But what does “soft law” actually mean in practice? Is it an accurate label for Japan’s current AI policy landscape?
Let’s explore the concept of soft law, how it applies (or doesn’t) to Japan’s AI guidelines and why this distinction matters. Rather than diving straight into legal jargon, we’ll take a step back and look at the broader picture: the intentions behind Japan’s regulatory style, the mechanisms it relies on and the implications for businesses.
1. What Is “soft law,” really?
Before we can assess whether Japan’s AI governance truly qualifies as “soft law,” it is worth unpacking what the term actually means.
Soft law typically refers to non-binding guidelines, principles or frameworks that influence behavior without the force of formal legislation. Unlike hard law — statutes and regulations that carry legal penalties — soft law relies on voluntary compliance, social norms and reputational incentives. It is widely applicable across various domains, especially in contexts where flexibility and adaptability are essential.
Soft law is particularly valuable in fast-evolving fields like technology, where rigid regulations can quickly become obsolete. For instance, when technologies such as digital platforms or artificial intelligence are newly introduced to the market, formal legislation often lags behind innovation. In such cases, soft law serves as a practical interim solution, offering guidance without the delays of formal rulemaking. Beyond technology, soft law plays an important role in areas like international relations, environmental governance and corporate social responsibility, where flexible, non-binding frameworks enable cooperation and progress without the constraints of rigid legal systems.
Japan’s recent legislative developments offer a timely illustration of soft law in action. The Act on Promotion of Research, Development and Utilization of Artificial Intelligence-Related Technologies (“Japan AI Act”) was enacted on 4 June 2025 (see this post reporting on the AI Act bill). While the Act encourages cooperation from businesses, it does not impose penalties for non-compliance. AI-related policies in Japan are typically issued as government guidelines or industry codes of conduct. Although these instruments are not legally binding, they exert substantial influence over expectations and practices across sectors.
This raises a critical question: how “soft” are these rules in practice? Do we label them “soft law” because they genuinely lack enforceability, or does the term obscure the real-world consequences of non-compliance?
2. Is Japan’s AI governance truly “soft”?
Generally speaking, it is fair to describe Japan’s AI governance as soft law. Except for the Japan AI Act, Japan does not have statutory laws specifically governing AI. Instead, a series of non-binding guidelines have been issued by government authorities, including the following:
- AI Guidelines for Business (Ministry of Internal Affairs and Communications (MIAC) and Ministry of Economy, Trade and Industry (METI))
- The Guideline for Japanese Governments’ Procurements and Utilizations of Generative AI for the sake of Evolution and Innovation of Public Administration (Digital Agency)
- Guideline for the Use of Generative AI in Primary and Secondary Education (Ministry of Education, Culture, Sports, Science and Technology)
- General Understanding on AI and Copyright in Japan (Agency for Cultural Affairs)
- Guidelines on Contracts for AI and Data in the Agriculture Sector to Protect Agricultural Know-How and Promotion of Data Utilization (Ministry of Agriculture, Forestry and Fisheries)
Although backed by strong institutional support, regular updates, and growing expectations for industry compliance, these guidelines are not binding laws. However, the absence of legal obligation does not automatically render a measure a mere suggestion. Non-compliance may still result in adverse consequences for businesses.
Take, for example, the Guidelines on Contracts for AI and Data in the Agriculture Sector to Protect Agricultural Know-How and Promotion of Data Utilization published by the Ministry of Agriculture, Forestry and Fisheries. These guidelines are not merely advisory. In fact, when farmers or agricultural cooperatives adopt smart farming equipment through government subsidies, compliance with these guidelines becomes a formal requirement in system service contracts.
This means that while the guidelines themselves may not be codified as law, they function as de facto conditions for accessing public support. In such cases, the line between soft law and enforceable policy begins to blur. The guidelines shape behavior not just through moral suasion or industry consensus, but through contractual obligations tied to public funding.
So, can we still call this “soft,” or is it something between soft and hard law — what some scholars call “semi-hard law”? Is it better understood as a policy-backed, incentivized framework — a hybrid model where voluntary principles are enforced through administrative leverage? The answer may depend on how these guidelines evolve and whether they begin to resemble enforceable standards in practice.
3. Why does Japan favor a soft approach to AI regulation?
Japan’s preference for soft law in AI governance is not accidental and reflects deeper cultural, political and economic considerations.
Culturally, Japan’s communication style sometimes tends to value flexibility and a degree of ambiguity. Rather than relying on direct or prescriptive messaging, policymakers may favor guidelines that leave room for interpretation and adaptation. This approach can help accommodate diverse stakeholders and evolving technologies while fostering voluntary compliance and collaborative engagement.
Politically, soft law enables the government to respond quickly to emerging technologies without waiting for formal legislation. It also allows different ministries — such as METI and MIAC — to work together and coordinate their efforts.
Economically, Japan aims to foster innovation while minimizing regulatory friction. By avoiding overly prescriptive rules, the government hopes to create an environment where AI development can thrive — especially among startups and SMEs — while still promoting ethical and responsible use.
4. Why soft law demands careful attention
Japan’s AI governance is often described as soft law, but as we have seen, this label can be misleading. The boundaries between voluntary guidelines and enforceable requirements are not always clear-cut. In some cases — such as the Ministry of Agriculture’s contractual guidelines — compliance is a formal condition for receiving public support or subsidies. Here, “soft” rules carry very real weight.
For businesses, developers and stakeholders, it is essential not to take the term “soft law” at face value. Always check what a guideline actually requires, how it is implemented and what non-compliance may entail. Even when a rule is technically non-binding, failing to follow it can mean losing access to funding, partnerships or market opportunities — and may carry reputational risks. In a rapidly evolving field like AI, staying informed and proactive is key. Soft law may offer flexibility, but it also demands responsibility and vigilance. Understanding the true nature of these guidelines — and the consequences of ignoring them — is not just good practice. It is essential for anyone navigating Japan’s AI landscape.
