Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsSome links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Japanese police arrested a 31-year-old Sapporo man in January 2026 on suspicion of displaying obscene electronic images after investigators alleged that he used generative AI to create more than 500,000 sexually explicit images resembling about 300 female celebrities.
Reports say the images were offered through a members-only website and that customers could request images of particular celebrities. The alleged operation generated approximately ¥11 million. The figure of 500,000—or approximately 520,000 in some reports—describes an investigative estimate of images created, not proof that every image was individually sold.
What police allege
The suspect, identified in English-language reports as either Tetsuro Chiba or Tatsuro Chiba, is a 31-year-old resident of Sapporo, Hokkaido. Because available reports use different romanizations, “Chiba” is the safest identification until an official Japanese record confirms the spelling.
According to police information reported by Tokyo Reporter and the South China Morning Post, Chiba began using a generative-AI tool around summer 2023 with images of approximately 300 people, including female idols. Investigators believe he produced more than 500,000 fake obscene images and made them available online between December 2024 and May 2025.
#1 Best Overall
Reports describe a paid-access model involving packages of 14 images, promotion through social media and additional commissions for requested celebrities. Police reportedly estimate that the activity produced about ¥11 million, roughly US$70,000 at the exchange rate used in the coverage. The suspect was reportedly quoted as saying he used the money for travel, entertainment or following celebrities.
These remain allegations. An arrest is not a conviction, and the available reporting does not establish the final charge, the number of customers or whether prosecutors will indict.
Why “selling 500,000 images” is imprecise
The headline figure needs careful interpretation. Reports variously describe more than 500,000 images or approximately 520,000. Those numbers were attributed to police sources and media reporting, rather than to a publicly available indictment or forensic inventory.
The evidence described so far supports an allegation that the suspect created a very large number of images and sold access to image packages and custom requests. It does not prove that 500,000 images were each sold separately, viewed by customers or associated equally with all 300 people.
The more accurate description is that police believe he created more than 500,000 sexually explicit images resembling hundreds of celebrities and commercially distributed access to some of that material.
What “AI deepfake” means here
A sexual deepfake is fabricated media that uses a real person’s face, likeness or other identifying features to make it appear that the person is nude or involved in a sexual act. The underlying event did not happen, but the false image can still cause serious reputational, emotional and professional harm.
The available reports do not establish the precise technical workflow. “Generative AI” could refer to synthetic image generation, face-swapping, image-to-image editing or another form of compositing. There is no confirmed evidence in the cited reports that Chiba trained his own model, used a particular commercial platform or created a specific type of fine-tuned system.
The celebrities have not been publicly identified in the available coverage. Naming them or linking to the alleged material would amplify its distribution and is unnecessary to understand the case.
What offense was reported?
The reported arrest allegation concerned displaying or making obscene electronic images available online. It was not described as a standalone criminal charge called “AI deepfake pornography.” Japan’s existing Penal Code framework includes provisions addressing the distribution or public display of obscene materials, including electronic records transmitted through telecommunications. Background materials are available from Japan’s Ministry of Justice.
That distinction matters. The alleged criminal conduct was not necessarily the private act of generating an image. Investigators appear to have focused on online display, paid access and distribution. The precise statutory provision and procedural status should be confirmed from police or prosecutorial records.
Japan’s Cabinet Office said in April 2025 that illegal AI-generated obscene material could be handled under existing laws on a case-by-case basis, including the Penal Code and child-pornography legislation where applicable. That position does not mean synthetic sexual imagery is consequence-free, nor does it amount to a blanket ban on every AI-generated sexual image.
Not automatically a revenge-porn case
Japan’s rules on the disclosure of private sexual-image records address certain genuine intimate images. The National Police Agency says public disclosure of qualifying private sexual images can carry a penalty of up to three years’ imprisonment or a fine of up to ¥500,000. Supplying material for that purpose can carry a penalty of up to one year or a fine of up to ¥300,000.
But it would be inaccurate to automatically call this incident revenge pornography. The available reports describe AI-generated images and do not establish that the suspect possessed or disclosed genuine private sexual images of the celebrities.
The categories overlap in harm but differ legally:
- Revenge pornography: disclosure of genuine private sexual material, typically without consent.
- Sexual deepfakes: fabricated sexual material using a real person’s likeness.
- Obscenity: a separate legal concept concerning the public distribution or display of sexually explicit material.
- Privacy, portrait and publicity rights: possible civil or other claims involving a person’s identity, likeness, reputation or private life.
The legal analysis may also change substantially when the depicted person is a minor, when genuine private material is involved, or when content is commissioned and commercially distributed.
Reported timeline
| Date | Reported development |
|---|---|
| Around summer 2023 | Chiba allegedly began using a generative-AI tool with images of approximately 300 people. |
| December 2024–May 2025 | He allegedly made obscene images available online through paid packages and requests. |
| By the arrest | Police reportedly attributed more than 500,000 images and approximately ¥11 million in proceeds to the operation. |
| January 2026 | Japanese media reported his arrest on suspicion of displaying obscene electronic images. |
Reports published around January 20–21 differ on the exact arrest date. Unless an official record confirms it, “arrested in January 2026” is the most reliable formulation.
Free tools Windows power users keep installed
One-click scans. No signup required.
Why the case matters
The alleged scale illustrates how generative AI can reduce the cost and time required to produce abusive sexual imagery. Publicly available photographs can be repurposed without consent, while ordinary online tools, social-media promotion and subscription mechanics can support commercial distribution.
Celebrity status makes the case highly visible, but the underlying risk is not limited to public figures. Ordinary people, former partners, colleagues and minors can also be targeted. A fabricated image can spread rapidly through copying and reposting, making removal difficult even after the original account or website is taken down.
The case also highlights several legally distinct stages: creating an image, possessing it, displaying it publicly, distributing it for payment, promoting it, hosting it and commissioning a particular likeness. Different laws and evidentiary questions may apply to each stage.
The Japanese government has separately warned about the misuse of image-generation and editing functions to create sexual images from other people’s photographs. Its approach, based on applying existing laws to particular facts, leaves the outcome dependent on the content, the people depicted, the method of distribution and the evidence collected.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWhat victims can do
Anyone targeted by a sexual deepfake should act quickly, while avoiding further circulation of the material:
- Save URLs, usernames, timestamps, messages and screenshots where legally safe. Do not download or forward illegal material.
- Report the content through the platform’s non-consensual intimate-image, impersonation or harassment process.
- Request removal from the relevant platform, website operator, host or provider.
- Contact police or an appropriate victim-support service, particularly where threats, extortion, minors or repeated harassment are involved.
- Consider legal advice about privacy, defamation, portrait or publicity rights and formal takedown requests.
The National Police Agency warns that rapid online spreading can make complete removal difficult, which is why preserving evidence and reporting promptly matters.
What remains unclear
- The exact Japanese statutory charge.
- Whether the image count was closer to 500,000 or 520,000 and how it was calculated.
- The number of customers, views and individually paid transactions.
- The identities of the celebrities allegedly depicted.
- The specific AI tool and technical process used.
- Whether prosecutors will indict and which provisions they will rely on.
- Whether any civil claims have been filed by people whose likenesses appeared in the images.
For now, the defensible conclusion is narrow but significant: Japanese police arrested a Sapporo man accused of using generative AI to create and commercially distribute access to a vast collection of sexually explicit images resembling celebrities. The case appears to have been pursued under existing obscenity laws, not under a single comprehensive national offense labeled “deepfake pornography.”
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
