Disney and Adobe have entered into a multi-year agreement to co-develop custom generative artificial intelligence models tailored specifically to the media giant’s operations, with the stated goal of scaling AI tools across its theme parks, film and television studios and sports media businesses. The partnership, announced on Monday, marks one of the most ambitious enterprise deployments of generative AI in the entertainment sector and signals that Disney intends to build proprietary capabilities rather than relying on off-the-shelf models that are available to its competitors.
The collaboration will begin with efforts to accelerate project development timelines at Disney’s Experiences division, the unit responsible for its global theme parks, resorts and cruise lines, as well as concept exploration at StudioLAB, the innovation hub housed within Walt Disney Studios. Adobe will also work alongside Disney to deploy AI tools aimed at deepening engagement with sports fans at ESPN and bringing more of the network’s extensive content archive to audiences. Through Adobe’s Firefly Foundry, an enterprise platform that allows organizations to build and deploy custom AI models, the two companies will develop generative models trained on Disney’s intellectual property that are capable of producing production-ready outputs. Critically, these models will remain private and proprietary to Disney, a detail that matters enormously for a company whose value is inseparable from the integrity of its creative franchises.
Walt Disney Imagineering, the creative, design and research arm behind Disney Experiences, has already put early iterations of the technology to work. The organization has deployed a custom image generation model alongside a sketch-to-image tool capable of transforming hand-drawn sketches into fully rendered concept art. Kyle Laughlin, senior vice president of research and development technology and engineering at Disney, described the impact in practical terms, noting that projects at the company are typically discussed in terms of years of development, and that Adobe’s platform has compressed creative processes that once took months into hours and days. He explained that the tools allow teams to run through multiple iterations quickly, which helps reduce costly changes once construction of physical experiences is underway. Laughlin was careful to emphasize that the technology is in no way meant to replace roles at Imagineering, a reassurance that reflects the sensitivity surrounding AI’s effect on creative employment.
That sensitivity is not abstract. Earlier this month, reports emerged that Disney was considering cutting thousands of jobs and had asked some managers to target twenty percent in annual cost savings. Against that backdrop, the announcement of an AI partnership inevitably invites scrutiny about whether efficiency gains will translate into workforce reductions, even as executives frame the technology as a tool for augmentation rather than substitution. The distinction between accelerating creative work and displacing the people who do it is one that Disney will need to navigate carefully, particularly given the cultural significance of Imagineering as the engine behind some of the company’s most beloved attractions.
The strategic logic behind the deal is difficult to dispute. Generative AI models trained on a company’s own intellectual property can produce outputs that are stylistically consistent with its brands, reducing the risk of the generic or off-brand results that often plague general-purpose tools. Proprietary models also address a persistent concern among media companies: that feeding sensitive creative material into public AI systems could expose valuable intellectual property or erode competitive advantage. By building private models through Firefly Foundry, Disney retains control over both the training data and the outputs, a structure that aligns with the cautious approach many entertainment firms have taken toward generative AI.
At the same time, the partnership raises questions that will not be resolved quickly. How Disney measures the return on this investment remains unclear, particularly in the Experiences division where physical construction timelines and capital costs dominate budgets. The extent to which ESPN’s archive can be monetized through AI-assisted discovery and personalization is promising but unproven. And the broader debate over AI’s role in creative industries, including ongoing labor concerns among writers, animators and visual effects artists, will continue to shape public perception of such deals. Disney’s framing of the technology as a creative accelerant rather than a replacement will be tested by the practical decisions the company makes in the months and years ahead. What is certain is that the entertainment industry’s experiment with custom generative AI has moved from the pilot phase into something considerably more consequential.
