Valve announces changes to AI content policies on Steam

The advent of AI tools for writing, art generation and voice work have been controversial in the video game industry since the surge in proliferation of these tools over the last few years. Valve reportedly attempted to crack down on games “utilising AI tech” last year. Since then, it has announced a change in policy which will enable the company to “release the vast majority of games” that use AI tools.

Going forward, developers submitting their games onto the Steam marketplace will be required to disclose if AI tools were used to develop the game, and if so, in what way. For pre-generated AI works in a game (such as AI art, voice work or writing), developers must promise that the resultant work does not include “illegal or infringing content” or misleading marketing materials. If the game contains live-generated AI features (that is, the game uses AI tools while the game is running), developers must also describe what kinds of guardrails are in place to ensure that illegal content is not generated.

To ensure transparency, Valve will include much of a game’s AI disclosure statement on its Steam Store page if the game is approved. As such, players will know going in if a game uses AI tools, and to what extent, and can take that into account when making a purchasing decision. New tools will also be implemented on Steam for players to report infringing material in games using live-generated AI-generated content not caught by any guardrails put in place by developers. Notably, despite this more permissive attitude towards AI-generated content by Valve, Adult Only sexual content created with live-generated AI will not be approved, at least for now.

Such a change in attitude was probably inevitable. With AI tools infiltrating nearly every industry in both benign and malignant fashions, at some point a blanket ban was probably going to be unworkable. Steam already has enough of a reputation for hosting shovelware as it stands without developers exploiting this loophole to flood the marketplace with low-effort and low-quality AI-generated content which crowds out the games actually worth playing. Hopefully features such as tags identifying games prominently featuring AI tools and community reporting will at least enable players to filter them out or the algorithm to de-prioritise anything that didn’t have any actual effort put into it.

This all could be subject to change, mind. Valve ends its statement by advising that “we’ll continue to learn from the games being submitted to Steam, and the legal progress around AI, and will revisit this decision when necessary“. It will remain to be seen what the long-term impacts of such a change in policy will be, and if other major storefronts, such as the Epic Games store, will follow suit.