Chinese Student Prosecuted for AI-Generated Deepfake Sexual Abuse Material
A Chinese doctoral student at a Seoul university was prosecuted for creating over 1,100 deepfake sexual abuse images and videos using generative AI, synthesizing the faces of seven female laboratory colleagues. The student deliberately circumvented the AI system's safety mechanisms and stored the material on a personal laptop before discovery.
What happened
A 30-year-old Chinese doctoral student at a Seoul university used generative AI called Grok to create 1,141 deepfake sexual abuse images and videos by synthesizing the faces of seven female laboratory colleagues onto pornographic material between November 2025 and May 2026.
The student searched for methods to bypass Grok's safety mechanisms during the creation process. The deepfake material was stored on the student's personal laptop and discovered during an unrelated research data transfer.
Reported impact
Five of the seven victims reported experiencing ongoing fear, trauma, and psychological suffering from the incident, and stated they had no intention to accept a settlement or return compensation money offered by the defendant.
Index analysis
The case involved intentional misuse of generative AI technology to create non-consensual synthetic sexual abuse material, coupled with deliberate circumvention of the AI system's safety guardrails.
The perpetrator actively searched for ways to bypass Grok's safety features, demonstrating premeditated intent to overcome AI safety mechanisms designed to prevent such abuse.
Governance controls
The defendant's defense attorney argued that the created images were never transmitted to third parties or distributed through the internet or social media, and were not created for distribution purposes.
The Seoul Prosecutors' Office sought conviction under South Korean sexual violence laws, requesting a three-year prison sentence, notification of sex offender status, and mandatory completion of a sexual violence treatment program.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.