Japanese Porn Depixelated by Man using Deepfake

A man allegedly unblurred, using deepfake technology, pixelated pornographic images and videos of pornographic actors, which violated Japan’s obscenity law requiring images of genitalia to be obscured.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.3 Fraud, scams, and targeted manipulation

The perpetrator used machine learning software to gain a personal financial advantage by creating and selling non-consensual sexual imagery of adult actors.

Additional risk subdomains

• 1.2 Exposure to toxic content: The incident involved the creation and distribution of illegal obscene/pornographic content violating Japanese obscenity laws.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The risk and subsequent legal violations were caused by a human user intentionally deploying machine learning software post-deployment to bypass obscenity laws for financial gain.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The report describes the creation of AI-generated content such as deepfakes, which falls under Risk Level 3 due to transparency and disclosure obligations.

AI system and alleged parties

• AI system: machine learning software
• AI purpose: Decensoring; Deepfake Video Generation
• Behaviour type: Tool
• Alleged developer: unknown
• Alleged deployer: Masayuki Nakamoto
• Alleged harmed parties: Japanese pornographic actors

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Minor, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Minor, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Negligible
• Psychological: direct Minor, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: The report describes 1 incident causing total financial loss of $96000.

Directly caused: Average financial loss per incident: $96000 in illicit revenue generated by the perpetrator.

Inferred additional harm: The actors and production companies likely suffered unquantified financial losses due to copyright infringement and unauthorized distribution.

Malicious content

Reported: The report explicitly describes toxic or malicious content created or spread directly or indirectly by the incident.

Directly caused: Nomoto created and distributed over 10,000 decensored pornographic videos and photos online.

Civil rights

Reported: The report explicitly describes violations of human or civil rights caused directly or indirectly by the incident.

Directly caused: The actors' rights to control their own likeness and bodily representation were violated through non-consensual sexual depiction.

Privacy

Reported: The report explicitly describes privacy violations caused directly or indirectly by the incident.

Directly caused: The actors' privacy was violated by reconstructing and depicting their genitalia without consent, although Japanese law did not have a specific charge for this.

Psychological

Reported: The report explicitly describes psychological harm, referencing 'online sex violence' and victims fighting to remove compromising fake videos.

Directly caused: The actors in the videos suffered distress and psychological harm from having their likenesses non-consensually decensored and sold online.

Inferred additional harm: It is highly likely that many of the actors whose videos were among the 10,000 manipulated suffered significant distress, anxiety, and reputational damage, though specific numbers are not detailed.

People affected

• Occurrences reported: 1
• People reportedly harmed: 10
• People reportedly exposed: 10

Potential causes

Management

• Unregulated Personal Site: Suspect managed his website without any external oversight or safety controls.

Technology

• ML Image Reconstruction: ML software was used to unblur and reconstruct pixelated adult videos.
• Accessible Deepfake Tools: Powerful deepfake-like tools are easily accessible to individual users.

Data Inputs

• Uncensored Training Data: The AI model used a large dataset of uncensored nudes to learn reconstruction.

Human Factors

• Financial Motivation: The suspect created and sold the doctored videos solely for monetary gain.

Process and Methods

• Direct Online Sales: Suspect sold over 10,000 manipulated videos directly through his own website.

Regulatory Environment

• No Specific AI Laws: Japan lacked laws criminalizing the use of AI to generate deepfakes.
• Inadequate Privacy Laws: The suspect could not be charged with violating the actors' privacy.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly outlines the facts of the arrest, the technology used (machine learning software for unblurring/deepfakes), the financial gains, and the legal charges. There is little ambiguity regarding the events.

A Japanese citizen was arrested for using machine learning software to decensor pixelated adult videos, generating $96,000 in illicit sales. While the incident represents a novel legal challenge and a violation of individual privacy rights, it was successfully resolved by local law enforcement and presents minimal implications for broader national security.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Japan
• Alleged perpetrator: Masayuki Nakamoto

Threat characteristics

• Imminence: Long-term. The perpetrator has been arrested and the active operation shut down, making this a long-term regulatory and legal concern rather than an active crisis.
• Autonomy: Human-controlled. The AI software was operated directly as a tool by the perpetrator to perform specific image reconstruction tasks without autonomous planning.
• Novelty: First-of-its-kind. Represents the first reported arrest in Japan involving the use of machine learning to reconstruct censored pornographic media, establishing a new legal precedent.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or human safety elements were targeted or affected in this incident.
• Information security: Negligible. The incident involved commercial obscenity violations and copyright infringement, with no connection to state-sponsored disinformation or intelligence compromise.
• Sovereignty: Negligible. No impact on government decision-making, state authority, electoral systems, or core constitutional processes.
• Economic security: Negligible. The financial impact was limited to illicit criminal revenue of approximately 11 million yen ($96,000), posing no threat to strategic industries or national economic security.
• Societal stability: Minor. Violated the privacy and likeness rights of adult actors through non-consensual decensoring. While harmful to the individuals involved, the impact was localized and managed through standard domestic law enforcement.