The rising use of artificial intelligence tools in the short video industry has sparked significant controversy, especially with recent incidents of celebrities having their likenesses used without permission in AI-generated short dramas.

On Monday, Hongguo Short Drama, a Chinese short-form drama platform under ByteDance, said that it had dealt with 670 short dramas in accordance with platform rules after a targeted review found they misused AI-generated materials. The platform said it had already taken down 1,718 comic-inspired short dramas that violated rules in the first quarter of 2026.

Hongguo is not alone. Once infringing videos are exposed, short video platforms often choose to remove them, as if this alone will resolve the issue. But such a response is far from enough.

In a landmark decision last month, Beijing's Internet Court addressed a portrait rights dispute that had captured public attention. The case involved a short video that used AI-generated face-swapping technology to create an image closely resembling a well-known actor, without permission. The court ruled against the producer of the video, citing unlawful use of deep synthetic techniques, and held both the producer and the broadcaster accountable for failing to meet their legal obligations.

This ruling underscores the law's importance in protecting personality and publicity rights. The law is very clear. Image rights protect an individual's interest against having their likeness disseminated or publicly displayed. Personality rights safeguard an individual's dignity, identity, and personal attributes such as image, name and voice. The right of publicity, which is primarily a property right, protects the commercial value of a persona and prohibits its exploitation.

But central to this protection is the "recognizability" criterion: if an unauthorized AI-generated image can be identified by the public as a specific individual, it constitutes a violation of their rights. The court's ruling reaffirms this principle, emphasizing that AI-generated virtual faces are not exempt from legal scrutiny.

The court's stance that "identifiability equals infringement" is a powerful reminder that both creators and distributors bear responsibility.

This begs the question: Why do AI-generated dramas that misuse celebrity likenesses continue to be broadcast?

Some image producers claim that the images are mere "technological coincidences", not deliberate acts of infringement. However, AI is a tool wielded by humans, and any misuse should be attributed to those behind its deployment.

This is not a failure of technology, but a failure of oversight, largely due to the low cost of infringement.

Nor is the practice limited to celebrities; ordinary individuals are also at risk. Once a video gains traction, even if it is removed, the creators will have already profited from the attention and can easily rebrand themselves to continue their activities. For those whose image has been used without their consent, the pursuit of compensation is costly and time-consuming. Celebrities may have the resources to fight back, but ordinary victims often struggle to make their voices heard.

To curb this practice, the legal repercussions for violators must be increased, and platforms must be held accountable for their prepublication review processes to intercept potential violations. Creating accessible channels for evidence submission could also empower both celebrities and ordinary individuals to better protect their rights.

Technology cannot be an excuse to infringe on a person's rights or break the law. Every face deserves legal protection. The entertainment industry and digital platforms must unite to safeguard the legitimate rights of not only celebrities but also those whose face may not be in the spotlight. It is imperative that the cost of violating legal and ethical standards be made high enough to act as a deterrent.