Opposing Steam’s AI disclosures, Epic CEO Tim Sweeney says game stores should drop “the AI tag” because “it makes no sense” when “AI will be involved in nearly all future production”

Echoing recent and divisive claims that “every game company is now using AI,” per Nexon CEO Junghun Lee, Epic CEO Tim Sweeney advocates for removing AI-generated content disclosures from stores like Steam and the Epic Game Store on the grounds that AI will be used in “nearly all” games.

Responding to a Twitter user who argued “Steam and all digital marketplaces need to drop the ‘Made with AI’ label,” Sweeney agreed. (And before we go any further, I do want to ask: if a game using gen AI is so great and normal, why do so many people seem keen on hiding its use from players?)

The Epic Games Store does not feature AI-generated content disclosures as a default, whereas PC gaming forerunner Steam has actually made them more visible in recent years. Arc Raiders, which has recently been in the crosshairs of this debate for its heavy use of AI voices, makes for an easy comparison. On Steam, Arc Raiders has the following blurb pinned to the bottom of its store page:

“The developers describe how their game uses AI-Generated Content like this: During the development process, we may use procedural- and AI-based tools to assist with content creation. In all such cases, the final product reflects the creativity and expression of our own development team.”

We’ve only just seen an outpouring of devs shutting down Lee’s claim that AI use should be treated as a default assumption in game development, and all of the reports from devs pushing AI-removed art and interrogating the labor implications of the tech can just as easily be aimed at Sweeney’s comments here.

Yes, a ton of large gaming companies and executives are investing in or talking up AI, from Ubisoft’s AI teammates and EA’s company-wide push to Take-Two’s bold claims on productivity and Gabe Newell’s predictions for an AI-heavy future, but equally a ton of teams actively avoid using the technology on principle, and are very vocal about that.

Weekly digests, tales from the communities you love, and more

(Image credit: Epic Games)

This is partly why Sweeney’s remarks can seem overzealous. “The rights situation” he mentions is central to ongoing discussions of gen AI use across disciplines, and perhaps especially in games where art and text may be scraped and repurposed without clear sourcing or permission. Multiple major game companies have come out to fend off potential unauthorized use of their IP by AI tools, for instance.

On the topic of store disclosures, it’s also worth revisiting the counterpoints in Valve‘s post about “AI Content on Steam” from January 2024, which laid out guidelines to make AI-aided games more transparent and, ultimately, releasable. “Under the Steam Distribution Agreement, you promise Valve that your game will not include illegal or infringing content, and that your game will be consistent with your marketing materials,” the company said. “In our pre-release review, we will evaluate the output of AI-generated content in your game the same way we evaluate all non-AI content – including a check that your game meets those promises.”

“Authorship disclosure” is a big reason stores like Steam make these gen AI blurbs so prominent. No matter how many executives suggest everyone is doing it and nobody cares, a sizable portion of people buying – and, again, making – games are demonstrably put off by gen AI and do want to know if and how it’s been used. Just look at the sour player response every time a game is caught using AI-generated assets without clear acknowledgement. Many people explicitly want real, human art made with observable intent, care, and methodology.

This whole discourse pocket feels like a microcosm of the AI wave in general: the biggest proponents, who often have the most to gain, treat an imagined AI future as today’s reality and incredulously question why everyone else hasn’t had the same, almost religious vision.

Microsoft shares study of gamers that reckons a whopping 79% are “open to help from AI,” but let’s all stop to consider the sample size: 1,500 people.