Industry executives at Filmart urged filmmakers to adopt hybrid AI workflows, showcased an AI-made feature and outlined how tools can free artists to focus on creative thinking
The Hong Kong film market known as Filmart became a focal point for discussion about generative AI when executives from broadcasters, studios and tech firms gathered to reframe the conversation. Instead of presenting the technology as a threat, the panel emphasized how AI tools can act as creative partners that accelerate production and expand visual possibilities. The session featured leaders who called for a shift away from old habits toward what they described as a more experimental, collaborative way of working.
Speakers included Lee Sangwook, head of the AI Content Lab at MBC C&I, Yuhang Cheng, COO of Midjourney China Lab, Fu Binxing, CEO of China Huace Film & TV, Ricky Lau, AI specialist lead at Google Hong Kong, and Yunan Zhang, vice president at MiniMax. The tone was largely optimistic: panelists framed current anxieties as a response to uncertainty rather than an intrinsic shortcoming of the technology. Throughout the discussion they stressed the need to unlearn entrenched production habits and to become fluent with a broader set of digital tools.
Among the items on display was an ambitious project from MBC C&I’s AI Contents Lab: an 80-minute feature titled Raphael, presented as an example of near-complete AI-generated filmmaking. Produced with a toolbox that included Midjourney and ElevenLabs, the film is directed by Yang Eekjun, Jung Juwon and Moon Shinwoo and blends science fiction, religion and action. Its story follows a war android modeled after the youthful likeness of an aging dictator who, while operating in a refugee district, crosses paths with a Catholic priest and begins an unexpected search for meaning. The project underscores how visualization tools can expedite concept-to-image workflows, and it is set to premiere in September.
The Lab presented a varied pipeline beyond the feature: short films, anthology mini-episodes and series that explore hybrid genres and AI-driven aesthetics. Titles include the six-episode drama The Fire Breakers, the magical fantasy Whispers Animal Clinic, the speculative Memory Proverb Chronicles, the AI art musical Alice in Joseon, and shorter pieces such as Art in the World 2: Judith, Sentiment and the micro-episodic Lua. The unit’s track record includes awards: the team behind Mateo won the Grand Prize at the 2026 Korea International AI Film Festival, and another title, Witness, earned the Grand Prize at the 2026 Busan International AI Film Festival along with a $10,000 prize. On the technical side, the Lab mixes tools like Stable Diffusion, Flux, Runway Gen3, Kling, Luma, VEO3, ElevenLabs, Suno AI and Udio, while using LLM-based development and Katalist AI for planning and storyboarding.
A recurring argument was that AI should serve artists rather than supplant them. Panelists suggested that fear around the technology often stems from not knowing how creative choices will be made when algorithms are involved. Yuhang described creation as the full spectrum of an artist’s emotional expression and maintained that those emotional decisions remain human. At the same time she highlighted that generative tools can act as rapid visualization engines: an idea that once required actors, sets and weeks of prep can now be represented quickly, allowing creators to iterate concepts and reallocate time toward craft and refinement.
Technical leads urged a pragmatic, mixed approach. Ricky Lau argued that productions will benefit from a hybrid approach that draws on multiple models and services, rather than betting on a single solution. He recommended that filmmakers become familiar with a wide pool of tools so they can choose the right mix for a given task. Fu Binxing predicted that within the next three to five years the barriers to creative production will drop dramatically, saying democratization of these technologies will enable even very young users to produce compelling work. The message was clear: adaptability and tool literacy are becoming part of a filmmaker’s skill set.
The conversations at Filmart suggested a practical roadmap: use generative AI to shorten repetitive tasks, free creative labor for higher-order decisions, and combine machine outputs with human judgment. Companies like MBC C&I are already translating that philosophy into experiments and commercial work, including branded spots and partnerships with platforms such as Wavve, and collaborations with advertisers like LG and convenience chain CU. For creators, the immediate invitation is to experiment responsibly with new workflows while preserving authorship and artistic intent.
Market attendees left with both examples and a directive: learn new tools, mix technologies thoughtfully, and rethink production processes that assume long lead times and heavy physical overhead. The industry debate is shifting from whether AI will be used to how it can be best integrated to augment creative teams. Filmart ends March 20.