AI Film and Television Competition Shifts Toward Premium Content: Zhixiang Yuanyi Bets on "AI + Live Action" and Professional Creation Tools
During the 2026 Golden Rooster and Hundred Flowers Film Festival, Zhixiang Yuanyi, together with Guomao Capital and Shanghai Film Co., Ltd., released the "New Film Creation Support Program," providing full-chain support from development and production to distribution and commercialization. DataEye-ADX data shows that in October 2026, new native AI drama/AI comic drama titles on Douyin reached 56,800, but new titles with fewer than 1 million views accounted for 80.31%, and the industry is shifting from overcapacity to competition over premium content. Its approaches include fully AI-simulated-human short dramas such as "The Bronze Strange Tales of Qinling," as well as an "AI + live action" solution that retains actors' performances while using AI to fill in scenes and visual effects.
Body · AI translation
According to the latest data from the DataEye-ADX industry edition, in October 2026 the number of new native AI drama/AI comic drama titles on Douyin reached 56,800, with an incremental playback volume of 280.849 billion, but new titles with fewer than 1 million plays accounted for 80.31%, while only 539 titles surpassed 100 million plays, accounting for 0.95%.
(1) After technology erased the threshold, the industry is undergoing a severe shakeout
The starting point of this round of industry reshuffling is not on the content side, but on the technology side. The spread of AI generation tools first erased the entry threshold for video production; making dramas is no longer the privilege of a small number of teams, and this directly detonated overcapacity on the supply side, which in turn triggered a severe chain reaction in the market.
Since the beginning of this year, after returns were extremely diluted, a large number of speculative players were forced to leave the market. At the same time, the demand side also cast a sober vote with playback volume: top-quality IPs have long occupied the top of the rankings, and homogenized new content is increasingly difficult to see.
Under pressure from the market, the incentive rules of mainstream content platforms have also tightened rapidly, not only strongly binding incentive income to reasonable per-drama duration, but also successively canceling some high heat incentives and series-scale bonuses.
From technology erasing the threshold and supply rapidly becoming excessive, to returns being diluted and companies exiting, to audiences turning to top-tier content and platforms stepping in to block padding, the entire causal chain points to the same result: the era of wild growth is over, and the industry must seek incremental growth from "premium content."
(2) Two breakthrough paths have already been proven
After squeezing out the water, what kind of content do users need?
Judging from the current distribution of genre tags in the market, counterattack, modern, female-oriented, urban daily life, male-oriented, drama, and rural themes are still dense, but reality-oriented, narrative-heavy themes such as urban daily life, drama, and rural are accelerating their push into the front ranks. When all teams can make the visuals exquisite, the marginal effect of visual spectacle declines rapidly, and the resonance of the subject matter and the quality of the script once again become the weights that determine survival.
Under this trend, two proven creative paths toward premiumization have gradually differentiated in the market:
The raising of content standards has set a higher threshold for AI generation technology: not only generating high-quality images, but also passing muster on character consistency, scene continuity, shot control, and multi-shot coordination, moving from "generating a single image" to "supporting complete series storytelling and professional production workflows."
Among the samples already proven, "The Bronze Mysteries of Qinling" illustrates exactly this point. In this fantasy suspense AI simulated-human short drama, across 60 episodes the character appearances, costumes and props, and scene art always follow the same set of specifications; in complex staging such as narrow mine tunnels and bronze mechanisms, there is no break in character blocking and environmental interaction. What AI has caught is no longer a single image, but the coherent narrative of an entire drama.
"The Bronze Mysteries of Qinling" topped Tencent Video's vertical-screen hot list just 12 hours after launch, and the second installment is about to launch on Tencent Video on November 22.
The second is aimed at higher-specification professional film and television production needs, exploring a creative approach that deeply integrates "live-action shooting and AI technology." This is based on real performances and live-action footage and pays more attention to "actors' emotional layers and the spatial texture of the set," which AI generation cannot replace.
The "AI + live action" approach fills in exactly the parts that are difficult to do in live action. It combines complex scene construction, visual effects, and shot creation, and refines the control of character states, spatial relationships, lighting details, and cinematic language to a deliverable and reusable level.
Data also validates this view. DataEye-ADX data shows that as of October 31, about 100 "AI + live action" short dramas were steadily airing each month, mainly concentrated in animal performance and live-action fantasy genres, with some titles exceeding 50 million in popularity on the Hongguo platform.
Although the production methods differ, both paths ultimately point to the same core, namely a leap in content quality.
When content standards rise, how can tools keep up?
The focus of market competition has shifted from pure generation efficiency to content quality. Then a new question follows: when the industry's standards for content quality rise across the board, how can AI film and television production capabilities keep up? What kind of tools do creators actually need?
(1) Creative control returns: letting technology accurately present creators' intentions
Many AI tools on the market today are accustomed to packaging a certain "hit style" into fixed templates, trying to let users "generate a finished film with one click." This approach appears to lower the difficulty of operation, but in fact leads to severe homogenization. For professional film and television production that requires fine storytelling, complex shot scheduling, and team collaboration, relying solely on random model generation and repeated "gacha" draws is clearly difficult to meet the production requirements of premium content.
Professional-grade AI film and television creation tools have chosen a different product path: rather than replacing professional creation with "one-click finished films," they break the core process of film and television creation into professional work stages that can be independently controlled and collaboratively completed, so that AI truly integrates into film and television teams' existing creative and production systems.
The underlying model's image generation capability and aesthetic understanding provide technical support for professional creation; refined workflow design allows these capabilities to be fully scheduled and controlled by creators. From script analysis, subject setting, storyboard design, and canvas creation to content generation, editing, and final delivery, each stage can be independently adjusted according to the director's creative intent and supports team members in completing work collaboratively according to professional division of labor.
Directors no longer need to accommodate generation results; instead, they can carry their own narrative judgment, visual style, and cinematic language through the entire production process, letting technical capabilities truly serve creative expression.
(2) Professional film and television solutions: letting AI deeply integrate into live-action production
For a high-quality film and television work, actors' performances have always been an important foundation for character building and narrative appeal. But in some current AI short dramas, problems such as mechanical character movements, exaggerated expressions, and distorted emotional expression still constrain the overall quality of the works.
Li Yaxian, head of the AI film and television business line at Zhixiang Yuanyi, mentioned that the actual pain points customers face in AI generation include issues such as “characters teleporting,” “movements violating common sense,” and “stiff expressions and distorted emotional expression,” and these clearly cannot be solved by image quality improvements alone.
In specific scenarios, in dialogue scenes, creators fully preserve the actors' emotions, tone, eye contact, and pauses, while AI adjusts the scene environment, lighting atmosphere, and character makeup and styling; in action scenes, the actors' physical reactions, movement trajectories, and performance rhythm are preserved, with AI used to complete scene replacement and special effects enhancement.
The value of this approach also extends to the “cleanup” of live-action footage. In actual film and television production, due to shooting schedule, location, and budget constraints, problems such as missed shots, scene effects falling short of expectations, and narrative connections requiring adjustment are not uncommon. Traditional methods often require reorganizing shoots, increasing production costs and timelines. “Live-action + AI,” however, allows more creative adjustments to be completed in post-production, further expanding the production space for live-action material.
Based on existing live-action material, creators can use AI to fill in missing shots, replace scenes, reconstruct lighting and visual style, and even adjust the narrative expression of original shots. The key is that these adjustments can continue the original character states, movement trajectories, and spatial relationships, allowing newly added footage to connect naturally with existing shots, rather than being regenerated detached from the original narrative.
This not only reduces repeated shooting and post-production rework, but also allows creators, on the basis of ensuring narrative coherence and visual texture, to further refine shot details and visual expression, so that improvements in production efficiency truly serve content quality.
(3) Upholding physical laws and narrative emotion
In previous models, characters “teleporting” was a failure of spatial causality; movements violating common sense were a failure of physical laws; and distorted emotional expression was the model's insufficient grasp of character states and narrative rhythm. The root causes all lie at the underlying level, and can only be solved at that level.
Who will safeguard good content?
Technology safeguards the lower limit of quality, creators determine the upper limit of content, and the support of industry and capital determines whether good content can go further and more steadily.
During the 2026 Golden Rooster and Hundred Flowers Film Festival, Zhixiang Yuanyi, together with Guomao Capital and Shanghai Film Co., Ltd., jointly released the “New Film Creation Support Program,” providing full-chain support from development and production to distribution and commercialization for selected projects that combine original value, innovative AI expression, and market potential.
Unlike content support programs that focus on support for only a single stage, this cooperation places greater emphasis on deep collaboration of industry resources. By linking technology, the film and television industry, and capital, AI technology no longer stops at tool application but enters the professional film and television production system, providing more complete industrial support for the sustained incubation and commercialization of high-quality content.
For an AI company that started with multimodal large models, moving from underlying models and professional creation tools further toward an industrial ecosystem layout not only expands the commercialization path of AI technology, but also means it is moving from a technology and tool supplier further into the development, production, and commercialization chain of film and television content, building deeper industrial competitiveness in the future film and television production system.
Of course, what will ultimately test this layout is not the speed of technological iteration, nor the scale of industrial resources, but whether technological and industrial advantages can be transformed into sustained productivity for high-quality content, driving AI film and television toward industrial-grade production standards and creating mature works that combine artistic value, content quality, and market competitiveness. This may be the key to AI film and television moving toward industrial maturity.