Epic Vows Developers Stay in Charge. Its Reported Orders Say Otherwise

Epic Games has confirmed a company-wide push toward AI-accelerated development, while reported instructions on a delayed Disney shooter put workers' control over their craft at stake. ^[1][3]

The unannounced game, reportedly featuring Marvel and Star Wars characters, had been scheduled for March 2027. According to Bloomberg's sources familiar with the project, as recounted by Eurogamer and GamesIndustry.biz on October 12, it has slipped without a new release date being communicated. ^[1][3]

The sharper news concerns how Epic reportedly wants the team to finish it. GamesIndustry.biz, citing Bloomberg, said developers were told on a video call the previous week to adopt an AI-first production process, including Anthropic's Claude, with staff overseeing its output. ^[3]

An internal message from a game director reportedly said, "No more hand-written code," and instructed that the art be "generated with full AI workflows." Those are instructions about who makes the work, not merely which software sits on a developer's computer. ^[3]

Epic offers a different emphasis. "We're embracing AI-accelerated development, led by Epic developers who are the visionaries and creative drivers for everything we do," communications director Jake Jones told Bloomberg. ^[1][3]

The company's promise deserves a fair hearing. It also deserves a hard test.

Creative control means having authority over how something is made, including when a tool is the wrong one. If the reported instructions hold, Epic is restricting that authority while assuring its workers that they remain in charge.

The game-specific details come from one Bloomberg report relayed by two publications, rather than two independent confirmations. The reporting does not establish that AI caused the delay, that this game's workers will lose their jobs, or that generated material will ship unchanged. Disney did not immediately respond to Bloomberg's request for comment. ^[1][3]

A tool becomes an order

There is a substantial difference between letting an artist use AI and telling an artist that all art must pass through it. The first gives a worker another method. The second makes the method a condition of the job.

That is why the reported wording matters. "No more hand-written code" leaves little room for a programmer to decide that writing a particular piece directly is the best way to solve it. An instruction to generate art through full AI workflows similarly places the production process above the maker's choice. ^[3]

Human oversight can be demanding, skilled work. A developer who checks generated code still needs to understand what it does. An artist who rejects an unsuitable image still needs an eye for composition, style and consistency. Reviewing a machine's output does not make those abilities disappear.

But oversight and authorship are different jobs. A worker may exercise judgement over material without having the same freedom to originate it. Epic's assertion that developers remain the creative drivers therefore needs more substance than a promise that humans will inspect the results.

Can they reject the workflow itself? Can they write code or make art directly when they judge that necessary? Will their judgement determine the production schedule, or will the schedule determine how much judgement they can exercise? Those are the questions that separate useful assistance from a compulsory change of profession.

Epic chief executive Tim Sweeney has described a broader reorganisation. In a post on X quoted by Eurogamer, he said a "huge pivot" was underway, with developers moving around for "maximum AI acceleration." He said the purpose was not "to do the same stuff with fewer people," but to let staff build "MORE and BETTER stuff" and bring the benefits to Unreal teams. ^[1]

More and better are two promises. Faster production might help deliver the first. The second depends on what developers are allowed to do with the time gained.

The promise after 1,000 job cuts

Epic announced in March 2026 that it was cutting 1,000 jobs. GamesIndustry.biz reported slowing engagement in Fortnite among the reasons. At the time, Sweeney said, "we're spending significantly more than we're making, and we have to make major cuts to keep the company funded." ^[1][3]

That history makes the company's current assurance about staffing harder to take on trust. It does not prove a concealed plan to replace workers. It does show that Epic has recently made a large reduction in its workforce under financial pressure.

A company can sincerely want its remaining staff to make more games and still change their jobs in ways they dislike. Headcount is only one measure of what workers stand to lose. The freedom to practise a craft, the credit attached to it and the scope to develop new skills matter too.

Sweeney's case is strongest when considered on its own terms. If AI reduces work developers find repetitive and gives them time for better ideas, the gain could be real. Rejecting that possibility would turn a labour argument into a rule against tools.

The reported mandate is the problem. It treats AI adoption as the starting requirement, before the public has been shown what the process improves. Workers are being asked to trust a conclusion that management has already reached.

GamesIndustry.biz also reported, citing Bloomberg, that the project fell behind partly because Epic underestimated the staffing and time it needed. Eurogamer's account places the AI-first instruction after the delay. ^[1][3]

If that account is accurate, Epic is changing the production method on a project whose original resource estimates were already wrong. That calls for particular care. A new workflow cannot by itself establish that the new schedule is realistic, or that the team has enough people to carry it out.

The test of Sweeney's promise should be practical: what tasks disappear, what responsibilities grow, and whether staff have the time and authority to make the extra work better. "More" is easy to ask for. "Better" requires room to say no.

The people closest to the work are sceptical

The 2026 State of the Game Industry survey gives the dispute a wider setting. Among more than 2,300 respondents, 36% said they used generative AI tools as part of their job. The share was 30% at game studios and 58% at publishing companies, support teams, and marketing or public-relations firms. ^[2]

AI is already part of work for a substantial minority of respondents. Yet use has not translated into broad confidence. Some 52% thought generative AI was having a negative effect on the industry, against 7% who thought its effect was positive. That is more than seven negative assessments for every positive one. ^[2]

Concern was particularly high in the disciplines touched by the reported Epic instructions. Negative views reached 64% among visual and technical art workers, 63% among game design and narrative workers, and 59% among programmers. ^[2]

These are survey respondents, not a census of the workforce or a measure of sentiment inside Epic. They do, however, show that the company is pushing into a production change many people in the relevant crafts regard as harmful. ^[2]

Calling that resistance fear of progress would be an inadequate answer. The people who make images, stories and systems have a direct interest in whether new tools improve those things. Their judgement should be part of the experiment.

Epic says the ambition is to build "MORE and BETTER stuff." ^[1]

The useful response to worker scepticism is to make that ambition measurable. Give teams meaningful discretion. Explain how review time counts in production estimates. Show how credit and compensation will work when a person's role shifts from making material to supervising its generation.

Those are demands for accountable management, not a ban on AI. A studio that believes its tools empower developers should be willing to demonstrate where the power sits.

Disney's investment raises the creative stakes

Disney invested $1.5 billion in Epic. GamesIndustry.biz reported that the delayed shooter was expected to be the first game in a partnership spanning multiple games. ^[1][3]

Disney's own account of the collaboration describes a persistent entertainment universe combining gaming, storytelling and creator expression. Its August 13 account of its games strategy emphasised original stories designed for games, rather than simply retelling or extending films and television series. ^[4]

That is an ambitious creative brief. Familiar characters can attract attention, but original game stories require decisions about what players do, how a world responds and why the experience is worth returning to. Disney's stated ambition makes the authority of the people making those decisions central to the partnership.

The strongest defence of Epic's approach is that players will judge the finished game. A satisfying game does not become unsatisfying merely because its developers used generated material. Human-made work also fails. This report cannot settle the quality of a game that has not been released.

But player enjoyment will not answer every labour question. A commercially successful result would not automatically justify stripping skilled workers of discretion, just as an unsuccessful one would not prove that every use of AI was mistaken.

My reading is that Epic's reported mandate puts its own promise at risk. Developers cannot be fully trusted as creative drivers while being denied the choice of how to make the work. The company should preserve that choice, then judge the tools by what teams actually achieve.

Epic and Disney's next public answers should therefore go beyond enthusiasm for acceleration. They should confirm the project's status, explain the production rules and say what authority developers retain. After that comes the harder evidence: the working conditions of the people making the game, and the game players finally get.