Generative AI remains one of the most divisive technologies in the video game industry, but a growing number of developers appear to be taking a cautious approach. According to video game lawyer Haley MacLean, anti-generative-AI clauses have become increasingly common in contracts between game developers and publishers, with virtually all of her clients now asking for restrictions on the technology.
MacLean, who heads the video game practice at Voyer Law, discussed the trend in an interview with GamesRadar+. Her work covers publishing agreements for indie and AA studios, giving her a direct view of how attitudes toward generative AI are changing within the business side of game development.
Developers Are Increasingly Blocking Generative AI
Speaking to GamesRadar+, MacLean said anti-AI language has expanded dramatically during the past year. She explained that clauses restricting generative AI were once primarily associated with more cautious publishers, but are now becoming increasingly common across smaller agreements as well.
“It’s turned around, especially in the past year,”
MacLean said. She added that anti-generative-AI provisions have effectively become boilerplate in many contracts, with developers and publishers seeking protection against potential problems involving AI-generated game assets.
MacLean also said that when a contract does not contain such restrictions, she may recommend adding them, including provisions covering AI use in areas beyond the game’s development. That can include marketing, quality assurance and porting work performed by publishers.
The lawyer’s comments echo concerns previously expressed by developers themselves. Pocketpair, the studio behind Palworld, has said it does not use generative AI because of concerns about how players would respond to it.
Copyright Risks Are Another Major Concern
MacLean’s opposition to unrestricted generative AI is not solely based on player sentiment. She also highlighted unresolved questions surrounding copyright ownership and potential infringement, arguing that studios could face legal complications when using AI-generated material.
One concern is whether an AI-generated asset can be protected as the studio’s intellectual property if there is insufficient human creative involvement. At the same time, companies could potentially face accusations that the generated material resembles copyrighted work used to train an AI system.
MacLean pointed to recent controversies involving AI-generated game assets as examples of why developers are becoming more cautious. She believes companies that use AI extensively could eventually face lawsuits as questions surrounding ownership and infringement become more prominent.
“Don’t touch it. It’s not worth the legal liability that it brings to you,”
MacLean said in the interview.
She also argued that companies could eventually spend so much time reviewing and modifying AI-generated content to establish sufficient human involvement that traditional human-created assets become the more practical option.
According to MacLean, the broader issue is that current copyright frameworks remain centered on human creativity. She emphasized that generative AI itself cannot own or transfer copyright, leaving studios to navigate an uncertain legal environment while the technology continues to develop.
