Japan: The AI Law With No Fines or Bans — How the "Name and Shame" Model Works

What happens if an AI law has no fines? Japan decided to find out: its first AI legislation relies entirely on public exposure as its enforcement mechanism, not economic sanctions.

A law built on pre-existing soft law

Most provisions of the AI Promotion Act took effect on June 4, 2025, formalizing the “soft law” approach Japan had already been applying on a non-binding basis — the Governance Guidelines for Implementation of AI Principles (2022) and the Guidelines for AI Business Operators (2024). The chapters establishing the AI Strategy Headquarters and the AI Basic Plan took effect later, on September 1, 2025.

No fines, no bans — the "name and shame" model

Unlike almost any other AI law in the world, the AI Promotion Act imposes no mandatory requirements, fines, or bans for non-compliance. Its enforcement mechanism is administrative guidance and a "public exposure" model: authorities can publicly identify AI business operators who don't cooperate, betting that reputational harm is incentive enough without needing an economic sanction.

The most recent guidelines: December 2025

The government published the Guidelines for Ensuring the Appropriateness of Research, Development, and Utilization of AI-Related Technology (AI Utilization Guidelines) on December 19, 2025, recommending a risk-based approach and comprehensive governance across the entire AI lifecycle — the substantive content keeps arriving via non-binding guidelines, not via the law itself.

Why Japan bet on this approach

The philosophy behind the law is explicitly "innovation first": the Japanese government prioritizes not slowing domestic AI development with a heavy regulatory burden, betting instead that multi-stakeholder cooperation and reputational incentives achieve the same governance goal without the compliance cost of a fines-based regime like Europe's.

What it means for companies operating in Japan

For a company accustomed to the EU AI Act's sanctions regime or even Singapore's lighter model, Japan is the most permissive extreme among the major AI economies — but "no fines" doesn't mean "no expectations." The non-binding guidelines remain the standard that regulators, Japanese enterprise customers, and business partners will expect you to follow, and the reputational risk of being publicly called out for not cooperating can be just as commercially costly as a fine in another jurisdiction.