Valve
Last week, we shared an anonymous report that Valve blocked at least some games from Steam that use AI-generated artwork. Over the weekend, Valve confirmed that report, telling Ars in an emailed statement that the company will block games that use AI-generated content unless developers can prove those AI models have been trained with data that doesn’t “infringe existing copyrights”. “.
“The introduction of AI can sometimes make it more difficult to demonstrate that a developer has sufficient rights to use AI to create assets, including images, text and music,” Valve spokesperson Kaci Boyle told Ars. “In particular, there is some legal uncertainty regarding data used to train AI models. It is the developer’s responsibility to ensure they have the proper rights to submit their game.”
Boyle emphasized in the statement that Valve’s “purpose is not to promote the use of [AI-generated content] on Steam” and that the company’s “priority, as always, is to try and ship as many of the titles we receive as possible.” Generative AI will “undoubtedly create new and exciting experiences in gaming,” Valve continued.
At the same time, the company says its hands are tied by the current state of the law. “Basically, our review process is a reflection of current copyright laws and policies, not an added layer of our opinion. As these laws and policies evolve over time, so will our process,” Valve said .
Far from settled
Despite Valve’s blunt claim that it doesn’t use “opinion” to interpret “current copyright laws and policies,” the copyright status of most AI models is far from certain. In large part, that’s because current copyright laws were written long before this kind of large-scale AI modeling was technically possible.
The companies using these AI models argue that machine learning based on copyrighted works falls under fair use, similar to human artists being influenced by the art they study, reference and remix. But a number of high-profile lawsuits filed by artists and stock art companies vehemently dispute that argument, saying these AI models wholesaled their content without permission.
Until those lawsuits generate some modern jurisprudence on the subject, the position of “current copyright law” around this content is far from clear. “I’m more concerned than ever about whether training is fair use in cases where AIs produce outputs that could compete with the inputs they’re trained on,” Cornell law scholar James Grimmelmann told Ars in April.
Given that uncertain legal environment, organizations such as Getty Images, Newgrounds and the scientific journal Nature have explicitly prohibited employees from using AI-generated art. At the same time, companies from Marvel to DeviantArt have embraced the use of the technology to some degree.
Valve is taking the more conservative route and avoiding what it calls “any legal uncertainty” by simply rejecting AI content trained on copyrighted material. And there’s nothing wrong with that. But in doing so, the company is effectively using its own interpretation of copyright law, even though it says the decision has no “extra layer of our opinion.”
