--- title: "Seven People Made a Feature Film: How AI Video Reached the Cinema" date: 2026-08-28 category: Generative Media site: NeuroAI canonical: https://neuroai.site/a/na-ai-video-cinema language: en --- # Seven People Made a Feature Film: How AI Video Reached the Cinema > A 100% AI-generated feature film made by a team of seven is heading to Korean theatres. Broadcast networks are using generative video in prime-time programmes. The tooling question has shifted from 'can it?' to 'who operates it? The moment generative video stops being a novelty is not when the clips get prettier. It is when a professional production company puts it on air and puts its name on the result. That happened in 2026. ## Key takeaways - **A feature film, seven people.** Working with MBC C&I, Korean studio **Mateo AI Studio** used Kling to produce *Raphael* — described as a **100% AI-generated theatrical feature**, about a robot's escape and spiritual journey, made by a **team of seven**, with a planned Korean theatrical release. - **Broadcast adoption.** **Studio Meta K** has deployed generative AI across well-known Korean programmes on KBS, MBC, ENA and MBN, using it in broadcast, advertising, drama and branded content. - **4K, natively.** In April 2026, Kling launched what was reported as the **industry's first native 4K direct output** in its video 3.0 series — no upscaling pipeline, aimed squarely at film and advertising professionals. - **The industry circuit.** Kling has presented at the **Busan International Film Festival**, **MIPCOM** in Cannes and **TIFFCOM** in Tokyo, and has begun recruiting local partners as authorised channel partners in Korea. - **Market trajectory.** The global text/image-to-video market has been projected to grow from about **$1.23 billion in 2025** to **$21.6 billion by 2034**, a compound rate of roughly 46%. ## Why a seven-person feature matters A conventional animated feature requires hundreds of people. Cutting that to seven does not mean the work disappeared — it means the work moved from **execution** to **direction**: prompting, curating, iterating, editing, and making aesthetic decisions at speed. This is the pattern that generative tools follow when they mature. They do not remove the need for taste; they reallocate labour away from frame-by-frame production and toward judgement. Studios that understand this early gain an advantage that is not about technology at all. The Korean case is instructive because it happened in a market with a sophisticated domestic film and television industry and no particular reason to favour a foreign tool. The tool won on output quality and workflow fit. ## What "professional grade" actually required Three technical problems had to be solved before professionals would touch generative video: **Consistency.** A character that changes face between shots is unusable. Chinese analysis specifically credits Kling's advantage in **frame consistency** as a reason for its business-market expansion. **Resolution and finishing.** Content that needs heavy post-processing is not faster than shooting. Native 4K output removes a pipeline stage, which is why it was framed as a cinema-grade capability. **Workflow integration.** Professionals do not want a demo; they want an API and an editing timeline. Kling 3.0's all-in-one design — text, image, audio and video natively in and out, with understanding, generation and editing in one flow — is a response to exactly that. ## The creator economy underneath The professional tier sits on top of a very large amateur base. Kling reported **over 100 million global users** and **more than 600 million videos generated** in two years, with monthly active users above **13 million** and growth rates that topped both outbound and global growth charts in May 2026. Volume of that kind is not vanity. It is a feedback engine: a hundred million users generate the usage signal, the failure cases and the stylistic demand that tell a model team what to fix next. Chinese analysts describe it as a flywheel where users, iteration and brand reinforce each other. Kling has also run creator competitions across Korea and other markets, working with universities and professional institutes to build the next generation of users. ## The open question Generative video has unresolved questions that no amount of quality solves: copyright and training data, disclosure obligations for AI-generated content, and the economics for the mid-tier freelance creatives whose work is most directly substitutable. But the technology question is settled. A seven-person team can make a feature film and get it into a theatre. The industry now has to answer the questions that follow — about credit, consent and who captures the value. *Based on company disclosures and Korean and Chinese media reporting in 2026; market projection from Intel Market Research.* --- Published by NeuroAI (https://neuroai.site/) — https://neuroai.site/a/na-ai-video-cinema Free to quote with attribution and a link to the original.