Valve explains its recent refusal to release an AI-based game


Recently, gamers have been regularly discovering AI-generated content from programs like Midjourney and DALL-E in many popular titles, including some on Steam. However, one developer has stated that they cannot release their product unless they can prove to Valve that they have ownership of the assets used to train the AI ​​algorithms.

User potterharry97 poured out his heart on the “aigamedev subreddit”, detailing his recent attempt to bring the game to Steam. This developer’s game contains some “apparently AI-generated” content. Valve customer support contacted the developer and informed that the AI ​​content needs to be removed before they are allowed to release their game on Valve’s platform.

If potterharry97 is unable to remove all AI content or can’t prove that he has the right to own it, Valve declares that it will not license or ban the game on Steam. The developer attempted to remove any indication of AI-generated content, but their game was rejected a second time.


A Valve representative shared the company’s stance on the use of AI assets when seeking permission to release games on Steam. “The use of AI can sometimes make it more difficult to prove that a developer has full ownership of AI-generated content, including images, text, and music,” the representative said. . More specifically, legal uncertainty is associated with the data used to train AI models. It is the developer’s responsibility to make sure they have the proper permissions to release their game.”

Valve’s goal is not to “prevent” the use of AI-generated assets on Steam, rather the company wants to look into how to integrate these games into its existing review policies. Two games using AI-generated images that have successfully landed on Steam are High on Life and Atomic Heart. Both games are likely to have developers proven to take ownership of the assets used to train the AI.

See also  Vietnamization of GTA 5 - super "popular" RPG for young people