From the Monkey King’s Romance to AI “Stars,” AI Is Rewriting China’s Screen Industry

Chongqing - AI tools are fueling a boom in fan remixes of classic TV, creating virtual actors with real fan bases, and raising hard questions for actors, studios and audiences.

Sun Wukong, the rebellious Monkey King, and Lin Daiyu, the tragic heroine of Dream of the Red Chamber, never meet in China's literary classics. Online, however, they have become one of the year's most talked-about couples.

An AI-generated image of the Monkey King and Lin Daiyu as a couple, created by an internet user.

Using AI tools, creators have spliced footage from the beloved 1986 TV adaptation of Journey to the West and the 1987 adaptation of Dream of the Red Chamber into new stories that pair the two characters. Fans see a shared rebellious streak in them: one defied heaven, while the other resisted the constraints of her world. Many viewers have embraced the pairing. For some, it offers the happy ending Lin Daiyu was denied, a long-standing regret among Chinese readers.

Not everyone is amused. Critics call the match absurd and worry that rewriting classic characters could blur cultural memory. Still, the trend shows how AI has made remixing classic films and TV dramas faster, cheaper and more widespread than ever. 

Earlier this year, a series of AI videos in the style of old Shaw Brothers martial arts films, in which a hero rescues a fox in the snow, spread across Chinese social media. Ordinary users, brands and even state media joined in with their own versions.

Commentators say the remix boom needs boundaries. Labeling a video "pure fiction," they argue, does not give creators license to distort a work's core spirit, its characters or actors' performances. Malicious parodies of heroes and gory or vulgar rewrites should be off-limits.

AI actors find real fans

AI is not just remixing old content. It is also creating new stars.

This summer, while many live-action TV dramas struggled, the AI-generated short drama The Girl Who Got Laid Off drew more than 200 million views. Its lead, Fang Taozi, is a small-town girl chasing a modeling career in the big city. Her story struck a chord with many young women.

AI actress Fang Taozi. (Photo from the internet)

Fang Taozi does not exist. Even so, her social media account, launched on June 14, has gained 557,000 followers. Some posts receive more than 700,000 likes. She shares fashion week photos, makes breakfast, works out and chats with viewers in the first person. She even interacts online with her on-screen love interest, and a separate "assistant" account posts behind-the-scenes content, much like a real celebrity's studio.

Fans copy her makeup and hairstyle and leave affectionate comments. Others find it unsettling to form an emotional attachment to an AI character. Some viewers have also accused the creators of blending facial features from real actresses, including Japan's Komatsu Nana. The director denied the claim.

A screenshot of AI actress Fang Taozi's Douyin page.

Fang is not the only one. Duan Yan, a white-haired male lead in AI-generated animated dramas, has become popular enough to sell merchandise. A bracelet his character gives the heroine sold more than 10,000 copies after merchants released a matching version. When a live-action remake of one of his dramas began filming, some viewers complained that the human actors couldn't match his looks.

Not every AI actor succeeds. In March, Youhug Media, the studio behind some hit dramas, signed two AI actors and faced an immediate online backlash over their uncanny look and resemblance to real stars. Months later, one of their accounts still had fewer than 10,000 followers. Fang Taozi's success suggests that audiences respond to a believable persona and everyday detail, not technology alone.

Why studios are betting on AI, and what could go wrong

For producers, the appeal is clear. AI actors never need rest, can appear in many productions at once, and cost almost nothing to reuse. They also never cause scandals. Britain's Particle6 says it created its AI actress Tilly Norwood in eight months, while a human actor typically needs three to four years of training before their first role.

The shift comes during a downturn for China's film and TV industry. Fewer projects are being made, and even established stars have reported receiving far fewer scripts. Short-drama actors have faced pay cuts and job losses. One popular actor reportedly cut his daily rate to a quarter of its former level and still struggled to find work. At one short-drama company, nearly 500 actors left within a single month.

Some veteran actors are choosing to license their likenesses instead. China’s Hong Kong actor Lawrence Ng Kai Wah has licensed his face from his 20s to an AI film. He said he enjoyed playing a leading man again without ever setting foot on set. Joey Wong, who retired from acting years ago, has licensed her on-screen image from her 20s and 30s to the online game Tianxia. Public opinion has softened: when veteran actor Tang Guoqiang endorsed an AI version of himself a few months ago, many people were puzzled. Now, some fans are asking studios to recreate other stars in their youth.

The risks are mounting, however. AI versions of real stars still often look stiff, and critics say the AI Joey Wong lacks the spirit of the real actress. Ordinary people face the threat of having their faces taken without permission. Voices are vulnerable too. Since 2023, clones of voice actor Shen Anyu's voice have spread across short video platforms, and his monthly income has fallen from 20,000 yuan (about 2,988 USD) to under 5,000. His own recordings are now sometimes flagged as AI.

Quality is another concern. According to the data platform Newrank, 11 AI short-drama accounts posted 1.34 million videos in a single month, mostly following the same revenge-fantasy formula. Critics say AI is being used more to cut costs than to tell better stories.

The legal picture is also unclear. Under China's copyright law, performers must be people or organizations. That means AI actors have no performers' rights and cannot sue if their work is copied or their image is attacked. A Shanghai lawyer has also warned that AI videos closely imitating Shaw Brothers films may infringe copyright even if the original works are old.

Overseas, the backlash has been even fiercer. After Tilly Norwood debuted last year, the US actors' union SAG-AFTRA said she was not an actor, and actors' unions in Britain and Canada also condemned her.

Opinions on what comes next are divided. Chinese actress Hao Lei has predicted that AI will replace 90 percent of actors, and said some AI performances are already better than human ones. Others point to unscripted moments that made film history, such as Jack Nicholson's improvised "Here's Johnny!" in The Shining, as proof of a spark machines cannot copy.

Many in the industry expect two tracks to emerge. AI is likely to take on low-budget, formulaic content and minor roles. Skilled actors who bring real-life experience to complex characters may become rarer and more valuable. Film did not kill theater, and television did not kill film. AI is reshaping the screen industry, but what lasts, many argue, will still be good stories and the human connections they create.