Skip to content
Submit incident
AII-1001-13179
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

AI-Generated Child Sexual Abuse Images Surge in Japan

Japan reported a surge in cases of child sexual abuse materials generated using artificial intelligence in 2025, with 114 cases involving images of minors created via AI, up from 110 the previous year. The victims were predominantly middle and high school students, with most perpetrators being schoolmates. The rise is part of a broader increase in online child sexual exploitation facilitated by social media platforms and online games.

26 February 2026 Β· Instagram Β· Harmful and abusive content
26 February 2026
AII-1001-13179View source β†—
Quick facts
Category
Harmful and abusive content
Country
Japan
Source
The Straits Times
Date
26 February 2026
Drafted by
Raidu Incident Detection

What happened

The National Police Agency reported that cases involving sexually manipulated images of children created using artificial intelligence rose to 114 in 2025, up from 110 the previous year, with middle and high school students accounting for nearly 90 per cent of the victims.

Students at the same schools as the victims made up the largest share of perpetrators, at 65. The report stated that many other victims are believed to have not yet consulted the police.

Reported impact

The rise in AI-generated child sexual abuse material is part of a broader surge in child sexual misconduct cases in Japan, with 4,858 cases in 2025, the highest level in the past decade. The number of children who became victims of nonconsensual sexual activity after meeting perpetrators through social media stood at 1,566, including 167 elementary school students.

Index analysis

Social media platforms and online games were cited as places where victims and perpetrators met. The most commonly used social media platform in such cases was Instagram, followed by X, with the Line messaging app serving as the most common method of communication.

Classification

Category
Harmful and abusive content
Industry
Media, social platforms and marketing
Country
Japan
Harm status
Actual harm
Intent
Intentional
Timing
Unknown
Harm types
Psychological, Human or fundamental rights
Harmed
Children
Failure modes
Deepfake image
AI systems
AI image generation systems

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

The Straits Times
Japan sees largest number of child sexual cases in decade amid online crime surge
26 February 2026