Steam opens the doors to games with AI-generated content and will mark its presence on the store page

Valve was updated Steam's content policy regarding the use of AI-generated content.

Valve launched last June 2023 block the release of certain video games on Steam, which used generative AI, as the developers did not own the copyright to the material they used for the game, although they denied having any intention of banning its use entirely.

At this point, they already indicated that they are working on updating the content review policy to adapt it to the use of generative AI, especially at the legal level.

Steam's new policy regarding the use of AI tools is to allow the publication of “the vast majority of games” that use them, as long as this is stated on the store page.

The content survey that developers must complete when submitting a game for release on Steam includes a section dedicated to AI and describes its use during the creation of the game and in the final product.

Valve distinguishes between two categories: pre-generated and real-time generated content.

In the first case, they include the generation of art, code, sound and others, as long as “they do not contain illegal content or violate the rights of third parties.” Game content and marketing materials also need to be consistent.

For content that is generated in real time (e.g. NPC responses), an additional requirement is added: explain “what measures you have taken to prevent the AI ​​from generating illegal content.” Players can have the presence Report illegal content in games that contain real-time AI-generated content from the Steam overlay.

Part of the AI ​​usage statement will appear on the game's Steam page so that “users understand how AI is used in the game.”

According to Valve, there will be an exception for the use of AI: “The publication of adult-only sexual content created using real-time generative AI” will be prohibited.

In recent months, legal pressure has increased on generative AI companies for training their models on copyrighted material without their creators' permission or compensation.

In a recent letter to the British House of Lords, the company OpenAI (ChatGPT, Stable Diffusion) openly acknowledged this his tools would not exist without access to copyrighted content.

Last December 2023 the newspaper The New York Times sued OpenAI and Microsoft for violating copyright law by using its content without permission or compensation. OpenAI has also received requests from a group of writers (including George RR Martin, John Grisham and Jonathan Franzen) and Getty Images.

Leave a Reply

Your email address will not be published. Required fields are marked *