China’s first fully AI-generated feature film, Sanxingdui: Future Past, has performed poorly at the box office, taking in roughly 1.7 million yuan — around £191,000 — despite a producer-estimated budget described only as “tens of millions.” The gulf between production outlay and ticket receipts has intensified scrutiny of whether AI can deliver commercially viable cinema at present.
A Guardian reviewer who attended a screening in Beijing reported the cinema was virtually empty and judged the film to be no worse than many conventional, effects-driven action pictures. That mixed assessment — noting technical novelty but limited audience appeal — mirrors wider questions about how viewers will respond to films made with generative tools rather than traditional human-led processes.
Industry reaction has been divided. Some investors and developers continue to place high value on AI-driven content creation, arguing the technology could lower costs, speed production and open creative possibilities once techniques mature. Others point to Sanxingdui’s weak box office as evidence that novelty alone may not be enough to draw audiences, and that storytelling, marketing and cultural resonance remain decisive.
The film’s performance also raises financial and strategic issues for producers who have already committed substantial funds to experimental AI projects. If returns fail to match investment, studios and financiers may grow more cautious, or they may shift to hybrid approaches that combine human creativity with AI assistance rather than complete automation.
For now, Sanxingdui: Future Past stands as a test case: a high-profile demonstration of what generative technology can produce on a feature scale, and a reminder of the gap between technical achievement and popular acceptance. Whether the experiment will spur further investment and innovation, or prompt a recalibration of expectations about AI’s role in mainstream filmmaking, remains to be seen.