Japanese police have arrested a 31-year-old man on suspicion of creating and selling sexually explicit images of female celebrities using generative artificial intelligence technology. The arrest was made by the Tokyo Metropolitan Police Department, which said the suspect uploaded deepfake content to paid membership websites and used publicly available AI tools to produce the material.
According to investigative sources, the suspect, identified as Tatsuro Chiba from Sapporo in Hokkaido prefecture, admitted to the allegations during questioning. Police believe he posted at least 2,200 explicit deepfake images to subscription sites between December 2024 and May 2025 and earned about 11 million yen (approximately US$70,000) from the activity. Authorities found some 520,000 stored deepfake images on his computer, many of which reportedly depicted fictionalised likenesses of actresses and pop idols.
Police allege the suspect used generative AI software to train models on images of around 300 female public figures, including actresses and entertainers, enabling the production of the explicit images. Investigators stated that the suspect promoted the content through social media to attract paying subscribers and offered custom image requests for higher fees. The deepfakes were hosted on membership platforms where users could access the material through paid subscriptions.
The case highlights emerging legal and enforcement challenges in Japan regarding the misuse of generative AI. Current Japanese law does not explicitly criminalise the mere creation of AI-generated deepfakes, but the distribution of obscene material and commercial exploitation can be subject to prosecution under obscenity and other statutes. Authorities report that this case represents a significant enforcement action against the commercial production and sharing of sexually explicit deepfake content.
Police investigations began after online reports of the deepfake content drew attention, prompting a probe into the suspect’s activities and the membership websites involved. The arrest was made on suspicion of distributing obscene images in violation of Japanese law, although charges and potential penalties have not been formally detailed publicly by prosecutors. Authorities say the investigation remains active as they analyse the scope of the deepfake production and distribution.
Experts have noted that generative AI tools have lowered technical barriers to creating realistic synthetic imagery, raising concern among legal scholars, law enforcement and civil society advocates. Japan’s legal framework is being scrutinised as officials and policymakers consider whether updates or new legislation are needed to address harms posed by non-consensual deepfake imagery and commercial exploitation of artificial intelligence technologies.
The Tokyo Metropolitan Police Department’s action forms part of broader efforts to tackle online offences involving AI misuse and the distribution of obscene digital content. Similar cases have emerged globally as authorities work to balance the regulation of digital technologies with the enforcement of existing obscenity and distribution laws.