TLDR: A 31-year-old man in Japan has been arrested for allegedly creating and distributing approximately 20,000 fake sexual images of 262 female celebrities using generative AI, earning 1.2 million yen. This marks the first time Japanese police have taken action against sexual deepfakes of celebrities created with generative AI.
Tokyo police have announced the arrest of Hiroya Yokoi, a 31-year-old resident of Akita city, in what marks a significant crackdown on the misuse of artificial intelligence in Japan. Yokoi is accused of creating and publicly displaying an estimated 20,000 sexually explicit deepfake images featuring 262 female celebrities, including actors, television personalities, and idols, utilizing generative AI technology. The alleged illicit operation generated approximately 1.2 million yen (around $8,000 USD) in sales between October of last year and September of this year.
Authorities stated that Yokoi admitted to his actions, explaining that he began producing the deepfakes to supplement his income, which he used to cover living expenses and repay a student loan. His method involved using readily available free generative AI software, with knowledge acquired from online articles and videos. He would feed images of celebrities into the AI program to generate the fabricated content.
The deepfake images were then distributed online to users who subscribed to a monthly fee. A premium subscription tier reportedly offered members the ability to request specific celebrities and poses for the custom-made fake images. The arrest specifically pertains to Yokoi’s alleged display of fake sexual images of three actors online across three separate instances between January 7 and June 2.
Also Read:
- Generative AI Propels Sophisticated Academic Fraud in Chinese Paper Mills
- Report Reveals AI-Generated Imagery Intensifying Anti-Muslim Hate in India
This case is particularly noteworthy as it represents the first instance where Japanese police have taken legal action against sexual deepfake images of celebrities created using generative artificial intelligence. The incident highlights a growing societal concern regarding the proliferation of deepfake technology. National Police Agency data from last year revealed over 100 identified sexual deepfake cases involving young victims nationwide, with 17 of these cases utilizing generative AI, predominantly created by classmates. This arrest underscores the escalating challenges law enforcement faces in regulating and prosecuting crimes involving advanced AI technologies.
