Epic Delays Disney Shooter as AI Takes a Bigger Development Role

Epic Games is reportedly delaying its Disney shooter while turning to AI-generated code and artwork to speed up production. The Unreal Engine developer has postponed the upcoming game featuring Marvel and Star Wars characters, originally targeting a March 2027 release, though no revised date has been announced. The report suggests that the project's slower-than-expected progress, partly due to Epic underestimating staffing and time requirements, is prompting the move toward an artificial intelligence-first approach.

An internal message obtained by Bloomberg revealed that a director instructed developers to "No more hand-written code," with the game's artwork to be generated using AI workflows. Developers were directed to utilize tools like Anthropic’s Claude, with human oversight of the results. Epic has confirmed its broader AI push, with CEO Tim Sweeney highlighting a strategic shift on X, stating the company aims to enable everyone to build more and better content.

The shift comes as Epic announced over 1,000 layoffs in March due to declining Fortnite engagement and spending exceeding revenue. While AI-generated code and art could potentially accelerate development, the project raises questions about software quality, development costs, and creative control. Developers must still test systems, fix errors, maintain consistency, and ensure components work together, even with faster generation.

The shift also impacts creative control, as recognizable characters from Marvel and Star Wars need to meet expectations for visual consistency, storytelling, and gameplay. The move could set a precedent for other studios, but the project's success or failure in quality and delivery time could either reinforce or dispel concerns about over-relying on generative tools.

While some industry professionals use AI for research and brainstorming, only 47% employ it for code assistance and 35% for prototyping. For development managers, it's crucial to measure time spent reviewing, integrating, and testing AI-generated output alongside time saved creating it, including costs associated with subscriptions, usage charges, and staff time to correct or maintain generated code.

Tracking total delivery time, defects, and rework can help determine whether AI improves productivity or shifts effort into later stages. Until Epic demonstrates the benefits with concrete data, software teams should view the approach as an experiment rather than a proven productivity model.