Deepfake Law in Japan
Japan regulates deepfakes through the Act on the Protection of Personal Information, defamation law, and the 2024 guidelines on AI-generated content. Relatively permissive compared to other major economies.
- Status
- enacted
- Jurisdiction
- Japan
- Effective
- Apr 2024
- Statute
- APPI + Penal Code + METI AI Guidelines 2024
Japan's deepfake regulation leans toward soft-law guidance rather than criminal prohibition. The AI Guidelines for Business published by METI in 2024 set expectations for AI operators; criminal prosecution uses existing Penal Code provisions on defamation and privacy.
Key provisions
METI AI Guidelines for Business (2024). Non-binding but authoritative guidance on responsible AI development and use. Includes recommendations on:
- Disclosure of AI-generated content.
- Consent for use of likeness or voice.
- Respect for human dignity and rights.
Non-compliance is not criminal but can inform administrative action and reputational exposure.
APPI (Act on the Protection of Personal Information). Biometric data provisions apply. Unauthorized use of voice or facial features for deepfake creation can trigger APPI obligations and administrative penalties.
Penal Code — Defamation (Article 230) and Insult (Article 231). Deepfakes that damage reputation or attribute false statements are prosecutable. Penalties include fines and imprisonment up to three years for aggravated defamation.
Copyright Act. Unauthorized use of a performer's voice or likeness may implicate right-of-publicity and neighboring rights. Japan has relatively strong protection for performers' rights.
Election Law. Paid political advertising must be identified as such; AI-generated content in campaigns falls under existing labeling obligations.
PPC enforcement
The Personal Information Protection Commission (PPC) enforces APPI. Has issued guidance on generative AI and data protection. Enforcement activity on deepfake-specific cases is limited; most enforcement is through criminal prosecution by police and prosecutors.
Dedicated deepfake legislation
Japan's AI Promotion Act (2025) set a broad governance framework but deliberately avoided heavy-handed restrictions to preserve innovation. Dedicated deepfake criminal provisions have been debated but not enacted.
Practical implications
For organizations operating in Japan:
- AI service providers: guidelines compliance is expected; formal licensing not required.
- Platforms: moderate takedown expectations; less strict than EU or UK.
- Enterprises: lower compliance burden than most peer economies; but defamation and copyright exposure remain real.
Japan is often considered a "wait and see" jurisdiction on AI regulation — less prescriptive than EU, China, or UK; more predictable than lightly-regulated Southeast Asian peers.