Disney has signed a multiyear deal with Adobe to build its own generative AI models, the kind of software that produces images or video from a text prompt, for use in Disney's production pipeline.
The partnership was announced at Adobe's MAX conference. The two companies describe the models as "artist-led" and proprietary, meaning they belong to Disney rather than being a shared public tool. Adobe will give Disney early access to its latest technologies, and the work is meant to reach well beyond film: theme parks, the cruise line and ESPN content are all named. Through Adobe Firefly Foundry, Disney will work directly with Adobe on models tailored to Disney's properties. The two will also explore "emerging technologies, including responsibly developed AI," according to Adobe's press release.
The notable choice here is custom over off-the-shelf. A general-purpose image generator knows nothing special about Disney's characters, house style or rules about how they can appear. A model built around a single company's catalog can, in principle, be steered toward that material and kept inside it. That is a different proposition from handing artists a public tool and hoping the output stays on-brand.
The announcement, which comes from the companies' own marketing language, leaves the hard questions open. It does not say what the models will be trained on, how much the deal is worth, or what "artist-led" means when the work gets done. Those details will decide whether this is a studio-controlled assistant for its own artists or a way to produce more content with fewer of them.