AI’s China Box Office Test Falls Flat on Opening Weekend

By Daniel Okafor

Bona Film Group reported that Sanxingdui: Future Past, a feature generated with artificial intelligence (AI), grossed RMB1.21 million (about $181,000) in Chinese theaters over its opening weekend. Bona’s blog put the budget at 10 million yuan, about $1.5 million.

The film became the first AI feature granted the right to play on Chinese screens. By Bona’s figures, that made it a box-office flop. The gross landed far below the cost. Not a nuanced opening, though the production itself was no one-click novelty.

What did the two-year production actually cost?

Bona said the film was produced over two years by a team of 70 to 80 people. Its blog post said some creators bristled at the idea that the images came from a single click. Technical work continued through final revisions, and the movie remained technically “shooting” until one week before its theatrical release.

The source compared the budget with that of a small independent film and set it against wider claims that AI would cut costs and let filmmakers make “Avatar” for next to nothing. Two years, 70 to 80 workers and revisions through the final week still ended with a modest ticket haul. The arithmetic was less flattering.

The source found the resulting artistry hard to see on screen. An attendee told international outlet AFP that the AI “falls short” of a human-made, live-action production. That verdict matters because free phone content and a paid theatrical seat compete on different terms.

Why did nearly 200 million views not guarantee a larger audience?

Before the feature, Sanxingdui: Future Past existed as an AI microdrama series, meaning short episodic videos, that attracted nearly 200 million views. Bona therefore had reason to expect a reasonable theatrical audience. The series had already demonstrated substantial online reach.

The leap from brief episodes to a 100-minute feature introduced a constraint. A microdrama could run only a few minutes, while the feature required a content review process before theatrical screening. That process removed the speed to bring something new to market, and “Sanxingdui” had already spent two years in production.

Her forecast was more generous. The attendee told AFP: “In the future, just like films, maybe decades later, it would get to a level when people are all willing to pay to watch,” adding, “But for the moment, I think there would be very few who would.”

That distinction between consumption and purchase is the commercial test. AI-generated content is already ubiquitous in China. A recent report, unnamed in the source, estimated that nearly two-thirds of videos in the microdrama boom were wholly AI-generated. Plenty of supply does not automatically create ticket demand. The old film-business adage said viewers would not pay to see an actor in a theater if television showed them free. AI video could now be consumed on phones for free.

Bona Film Group has already started production on a second AI-generated feature, Unplanned Operation in the Cold Moon. The company has also received requests to screen the film in countries outside China. Neither the sequel nor those requests will erase the first film’s result. They will show whether the business model can travel beyond a domestic opening.

Can Gossip Goblin build a theatrical audience in the United States?

YouTuber and AI creator Gossip Goblin is watching the Chinese experiment. His fully AI-generated feature, Gods Don’t Give Gifts, opens in U.S. theaters this December, with international sales launching at the American Film Market next month.

It is the first such feature to receive a rating from the Motion Picture Association (MPA). Its team also intends to submit the film for eligibility at the Oscars’ Best Animated Feature category. Those milestones concern classification and eligibility, not audience appetite. They do not settle the box-office question.

Gossip Goblin has cultivated a large audience through numerous psychedelic, sprawling AI shorts, many set within a connected story universe. He hopes to follow the Markiplier and Iron Lung playbook: mobilize a dedicated fanbase and bring it into cinemas through a self-releasing strategy.

The wager is straightforward. A dedicated online audience will buy tickets for the kind of theatrical experience it can also find at home or on social feeds for free. Viewers in China and the United States expect cinemas to offer a level of cinematic quality they cannot get on their phones. December’s U.S. debut will test whether Gossip Goblin’s following can convert attention into tickets, and whether the “Sanxingdui” experiment travels to American audiences.