Google DeepMind Hosts AI Film Challenge for 15 Teams

Google DeepMind hosts a San Francisco AI filmmaking challenge where 15 teams create original short films using the same starting footage.

Share
Google DeepMind Hosts AI Film Challenge for 15 Teams

Fifteen teams of AI filmmakers have been selected to take part in an unusual creative challenge at Google DeepMind in San Francisco, with every group given the same footage to develop into an original short film.

The Gossip Goblin GenJam, scheduled for 20 October 2026, brings together filmmakers, AI developers and creative technology companies to explore how artificial intelligence can be used to tell different stories from identical starting material.

Rather than entering films created over several weeks or months, participating teams have just one evening to develop their ideas, produce their footage and present their finished work.

Fifteen teams begin with identical footage

Every team receives the same 15-second clip from Gods Don't Give Gifts, an AI-assisted production developed by Gossip Goblin, alongside reference material showing characters and landscapes from the film.

The challenge is to transform those assets into a new short film lasting no more than two minutes, with each team free to choose its own creative direction.

Some filmmakers may develop the original story further, while others could experiment with different genres, animation styles or entirely new interpretations of the characters.

The shared starting point makes the challenge particularly interesting because differences between the finished productions will reflect the creative decisions and technical approaches adopted by each team.

Which AI filmmaking tools are involved?

Participants can work with their own AI production systems, allowing filmmakers and developers to experiment with the technology they already use.

Teams without their own production setup have access to Showrunner, the creative platform associated with Gossip Goblin, including existing prompts and assets intended to support their work.

Selected teams may also receive credits from technology partners, providing access to additional generative AI models and production services.

The competition brings together groups of two to four people, including independent filmmakers, software developers and professionals creating the technology behind generative video.

By allowing different tools and workflows, the event offers an opportunity to compare how filmmakers approach similar creative problems using different systems.

How are the winning films chosen?

Each team is expected to screen its completed film during the event, giving participants and guests the opportunity to see how the original footage has been interpreted.

The audience will vote on the productions, with a judging panel also contributing to the final decisions.

First place carries a US$2,000 cash prize, tickets to the Los Angeles premiere of Gods Don't Give Gifts, the opportunity to feature within the Gossip Goblin universe and additional technology credits.

The second-place team will receive US$1,000, while third place carries a US$500 prize, with both also receiving additional benefits from the organisers and technology partners.

The event is hosted by 645 Ventures, Machine Cinema, Google DeepMind, Showrunner and Gossip Goblin, with support from other organisations working across AI technology and entertainment.

Why is the challenge different from other AI film competitions?

Most filmmaking competitions begin with creators developing their own stories and submitting completed work for consideration, but GenJam approaches the process differently.

By giving everyone identical footage and reference material, the organisers can explore how the same fictional world might develop through the ideas of different filmmakers.

The concept is connected to what 645 Ventures calls Iterative Media, an approach in which digital content can be extended, adapted and reinterpreted rather than existing only as one finished production.

For filmmakers, this presents an opportunity to experiment with characters and storytelling while working within the restrictions of a shared creative brief.

The challenge also highlights the role of human decision-making in AI-assisted filmmaking, with participants responsible for directing their chosen tools and deciding how their finished stories should unfold.

Where can audiences discover the finished films?

Audiences interested in the project can follow Machine Cinema for future GenJam events and updates about its AI filmmaking activities.

The official Gossip Goblin GenJam listing also provides details of the competition, its participating organisations and the creative brief.