Japan's Soft Law And Korea's Binding Act Are Now The Two Asian AI Regulation Models Everyone Else Will Copy

Asia Now Has Two Very Different AI Regulation Templates

Eleven months after its enactment, Japan's AI Promotion Act and South Korea's AI Basic Act have finished bedding down into concrete operating rules.

What looked in 2024 like two countries moving in parallel now looks very different. Tokyo has chosen the softest AI law of any major economy.

Seoul has gone the other way and built the strictest comprehensive framework outside the EU. Every other Asian regulator now has to pick a side.

Japan: No Fines, No Bans, Just Guidelines

Japan's AI Promotion Act passed on May 28, 2025, with most provisions in force from June 4, 2025. The AI Strategy Headquarters, chaired by the Prime Minister, held its first meeting on September 13, 2025, and is now drafting the 2026 AI Basic Plan.

The Act does something that has not really been tried at scale^✦: it governs AI without a single mandatory prohibition. There are no pre-launch checks, no prohibited use categories, and no fines. Non-compliance is handled through voluntary guidelines from METI and the Ministry of Internal Affairs and Communications, and a "name and shame" mechanism that publishes the names of operators who refuse to engage. Data use for training is also highly permissive, with no opt-out right for rights-holders, which is part of why Japan is now one of the friendliest training jurisdictions in the world.

Korea: Binding, Extraterritorial, And Focused On High-Impact AI

Korea's AI Basic Act, in force since January 22, 2026, is Asia's first comprehensive AI statute. The Enforcement Decree published on January 21 sets the operational rules and draws the high-impact AI line at the 10^26 FLOPS threshold for training compute^✦. Above that line, operators must carry out risk assessments, document user protection measures, and submit to enhanced supervisory attention from MSIT.

We have aimed at minimum regulations to build public trust while supporting domestic AI competitiveness.

The Act also reaches beyond Korean soil. Any AI system whose outputs affect Korean users triggers the regime, including the requirement to appoint a domestic representative if the provider has no local presence. That is the same extraterritorial logic the EU used, and it is already reshaping how American and Chinese platforms plan Korean launches.

What The Two Models Share, And Where They Diverge

Both frameworks are promotion-first. Both avoid the EU's prohibited-use list. Both lean on sector guidance for specifics.

But the core philosophical divide is clear. Japan treats AI law as a developmental scaffold and trusts firms to self-regulate under published guidelines. Korea treats AI law as a user-protection backstop with teeth, even if those teeth are relatively small by EU standards.

The practical effect on multinational counsel is that a single APAC AI compliance playbook is no longer sufficient.

Korea requires labelling of generative AI^✦ outputs under Article 31 of the Act, including deepfake labelling. Japan encourages it.

Korea imposes operator notification duties. Japan nudges them.

For any global model provider serving both markets, the minimum sensible posture is Korea-grade compliance, because a Japan-only build is then trivially portable upward.

Which Model Is The Rest Of Asia Copying

The early returns are revealing.

Thailand's AI Act, now in force, looks more Japanese than Korean, with soft-law emphasis. Singapore's AI Verify stays firmly voluntary and export-focused.

India's April 2026 labelling rules sit somewhere in between, mandatory for a narrow set of generative AI categories but without Korea's extraterritorial reach. Indonesia's draft regulation leans Japanese.

That leaves Korea largely alone in the region as a fully binding, extraterritorial regime, a position that will shape where global AI launches first and where they quietly delay.

Hot-Button Contrasts

1. Fines: Japan none, Korea up to KRW 30 million per violation.
2. Bans: Japan none, Korea none. Both differ from the EU.
3. Training data opt-out: Japan no, Korea not explicit, EU yes.
4. Extraterritorial reach: Japan no, Korea yes.
5. Mandatory labelling of generative AI output: Japan no, Korea yes.
6. Designated domestic representative required for foreign providers: Japan no, Korea yes.

Frequently Asked Questions

Does Japan's AI Promotion Act have any enforcement at all?

Yes, but not through fines. The mechanism is publication of non-cooperating operator names, plus the ability to withhold government procurement, grants, or sector guidance cooperation. The real teeth are reputational and commercial, not monetary.

What happens if a US AI firm does not appoint a Korean representative?

Under the Korean Enforcement Decree, failure to appoint a domestic representative is itself a punishable offence with an administrative fine up to KRW 30 million, and it can also trigger corrective orders that compound the initial penalty if ignored.

Which Asian jurisdictions look most like Japan so far?

Thailand's AI Act and Indonesia's draft AI regulation both lean soft-law and promotion-first. Singapore's AI Verify framework is even further down that axis, sitting as a voluntary assurance toolkit rather than a statute.

Does Korea's AI Basic Act apply to open-source model providers?

Yes in principle, because the Act applies to AI systems with high-impact effects on Korean users, regardless of distribution model. However, enforcement focus so far has been on commercial operators above the 10^26 FLOPS threshold, with open-weight^✦ smaller models facing lighter scrutiny.

Closing

Asian AI regulation is no longer a single conversation. It is two. Which model do you want your firm's products aligned to by default? Drop your take in the comments below.