Ash Koosha Debuts AI-Generated Film at Tribeca

Brothers Ash and Pooya Koosha premiered Dreams of Violets, a 75-minute fictional drama about Iran's January 2026 protests, at the Tribeca Film Festival on June 10, 2026. The Koosha brothers, Iranian-born and based in London and Menlo Park respectively, created the film under Fountain 0, an AI film studio. Ash serves as CEO and director; Pooya co-founded the company and produced. Per Variety and Hollywood Reporter, the Tribeca premiere marks the first time a fully AI-generated live-action feature has been accepted at a major film festival.

Fountain 0 reports the film was built in roughly three months for approximately $2,000. The AI tool stack, confirmed across multiple trade publications, includes: Kling AI for video generation; Anthropic's Claude for language and editing tasks; Google's Gemini and a tool called Nanobanana for research and imagery; and Fountain 0's own proprietary technology for blocking and frame accuracy. Tom Rogers, Fountain 0's executive chairman, has described the low-cost approach as a way to democratize high-quality filmmaking for independent creators, per CryptoBriefing.

The film dramatizes civilian resistance during the January 2026 Iranian protests, a period in which international organizations estimated the civilian death toll at over 7,000. All characters and images are fully AI-generated, based on journalistic reports, photographs, and eyewitness accounts. The Guardian attributes to Ash Koosha: "I would say 80% of it is a recreation of events that actually happened," and separately reports he deliberately avoided modeling characters on living people, quoting him: "Because of the security issue, it would not be safe for the characters to even remotely resemble someone."

Response after the June 10 premiere was sharply mixed. TheWrap describes the film as fractured vignettes and reports it landed "with a thud rather than a revolutionary bang," flagging the limits of current AI tools in rendering political trauma convincingly. The Financial Times was similarly critical, characterizing the result as slop. The Guardian highlighted the film's novelty and low-cost workflow as a democratizing force, while also noting audience and festival hesitancy toward fully AI-generated work.

Three indicators will shape how this story develops for AI and data practitioners: first, whether Fountain 0 discloses model versions, training-data provenance, and prompt engineering; second, how Tribeca and other major festivals develop explicit policies for AI-generated submissions; and third, whether character anonymization becomes an industry safety norm in politically sensitive AI productions. Rights-clearance procedures for the visual training data remain an open legal and ethical question not addressed in current coverage.