In 2024 the Jiading District Court of Shanghai concluded “Photographer Chen v. Shanghai Yi×× Network Technology AI Face-Swap,” the first Chinese judgment to hold that AI face-swap is neither fair use nor an original adaptation, drawing a bright red line for the booming AIGC (AI-generated content) industry.

Case at a Glance: A 10-Second Costume Clip Gets Swapped
Photographer Chen posted 13 short videos of a woman in ancient dress on his Douyin account “Photographer ××.” Each clip lasts only about ten seconds, but Chen invested heavy creativity in framing, camera movement, costume matching and choreography.
The defendant company launched the mini-program “Mou Yan.” With an AI video-synthesis algorithm it replaced the face in Chen’s clips with the user’s selfie, leaving scenery, shots, styling and movements almost intact. Users either watch a 30-second ad or pay a membership fee to download the “self-starring” costume clip.
After gathering evidence Chen sued, demanding the infringement stop and RMB 50,000 in damages. The court found that apart from facial features the accused videos were “substantially similar” to the originals, ordered the defendant to pay RMB 7,500 plus reasonable costs and stop the infringement.
Key Holdings: Three Defenses All Fail
(1) Original adaptation?
The court held that AI merely performs a local facial replacement; it does not create new “originality” in audiovisual expression and therefore is not an adapted work under copyright law.
(2) Fair use?
The defendant used the original videos as a traffic magnet, kept plot and pictures intact, pursued a clear commercial purpose and substantially substituted for the original, so it did not qualify as “appropriate quotation,” teaching, scientific research or any other fair-use scenario.
(3) Technology is neutral?
The platform actively supplied the source clips, the algorithm and a paid interface; it participated in and profited from the whole process—upload, synthesis, download—so “tool neutrality” is no shield.
Industry Pain Point: Algorithms “Cut Corners,” Creators Get Hurt
AI face-swap, AI dubbing, AI watermark-removal… the technical threshold keeps falling while the grey supply chain matures. Some platforms feed others’ hit material straight into their models; users supply a photo and “instant blockbusters” pop out. Consequences:
Original crea