A fully AI-generated feature film, Dreams of Violets, directed and produced by brothers Ash and Pooya Koosha, will premiere June 10 at the Tribeca Festival, according to Variety and The Hollywood Reporter. The 75-minute docudrama was produced in about three months for roughly $2,000, with no cameras or human actors; Deadline and Variety report every image and person on screen was created using AI tools and journalistic sources. The film dramatizes events around January 2026 protests in Iran, a conflict coverage Deadline and Variety say left at least 7,000 dead. Tribeca co-founder Jane Rosenthal is quoted praising the film's emotional immediacy (Variety, Hollywood Reporter). Director Ash Koosha is quoted saying he would have preferred a conventional production but lacked access to Iran and personnel (Hollywood Reporter). According to Fountain O, the project is the first full-length, AI-generated film accepted into a major festival.
What happened
Dreams of Violets, a feature-length film produced and directed by brothers Ash and Pooya Koosha, will have its world premiere on June 10 at the Tribeca Festival, per Variety and The Hollywood Reporter. Deadline and Variety report the film runs about 74-75 minutes and was produced in roughly three months for about $2,000. Deadline and Variety say every image and character in the film was generated using AI video tools and informed by journalistic reports, photographs and eyewitness accounts. The film is described as a fictionalized docudrama inspired by protests and a January 2026 massacre in Iran, which Deadline and Variety cite as events that left at least 7,000 dead.
Technical details
Per Variety and Deadline, the Koosha brothers used a mix of third-party models and in-house tooling during production, including Kling AI for video generation, Anthropic's Claude for language-related editing, Google's Gemini, and Nanobanana for imagery and research, alongside Fountain O's own technology for blocking and frame accuracy. The Hollywood Reporter and Variety both quote Ash Koosha saying he recorded initial voicing before altering it with AI tools. Those outlets report the production replaced traditional actors, sets and cameras with an AI-driven pipeline executed from Koosha's home in London.
Editorial analysis - technical context
Industry-pattern observations: Generative video pipelines combining multiple specialised models and project-specific tooling are increasingly used to synthesize coherent narrative footage. Projects that stitch image-generation, video synthesis, large-language-model editing, and bespoke blocking tools typically trade off photorealism, controllability, and production time; practitioners use model ensembles and human-in-the-loop prompts to manage continuity and motion. Low-budget experiments like this one accelerate iteration cycles by replacing expensive physical resources with compute and prompt engineering, but they also expose practitioners to challenges around temporal coherence, lip sync, and ethical provenance of training data.
Context and significance
Public reporting frames Dreams of Violets as notable both artistically and institutionally, Variety and Deadline report that Fountain O and other coverage describe it as the first full-length, AI-generated film accepted into a major festival. Tribeca co-founder Jane Rosenthal is quoted in multiple outlets praising the film as an example of using emerging technologies for "deeply human storytelling" (Variety, Hollywood Reporter). Coverage in CBC and other outlets also records substantial public backlash to the trailer, with many online commenters denouncing the film's use of AI. The combination of festival legitimacy, politically sensitive subject matter, and the film's production method has provoked debate over the cultural and ethical boundaries of synthetic media.
What to watch
For practitioners: track technical breakdowns or behind-the-scenes material that validate claims about tooling and workflow, and watch for festival programming statements clarifying selection criteria for AI-generated works. For policy and governance observers: note whether rights holders, union groups, or festival bodies issue positions on synthetic performers and attribution. For model developers: observe how multi-model pipelines address temporal consistency and voice synthesis at feature length; the Kooshas' reported toolchain may influence best-practice patterns for future generative-video productions.
Scoring Rationale #
The story marks a notable industry application-festival acceptance of a fully AI-generated feature-relevant to practitioners tracking generative-video workflows and cultural impact. It is important but not a technical paradigm shift.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.