In the wake of a recent controversy surrounding a game with AI-created assets, Valve has clarified its position on such games. A developer last week claimed that their title, which incorporated AI-generated content, was turned down by Valve on the grounds of proving the ownership of rights used in the data set that trained the AI to generate those assets. The increasing use of generative AI tools by developers trained on human-made art sourced from the internet has brought forth hitherto non-existent copyright issues.
Valve recently issued a statement to VGC saying that it does not intend to discourage the use of AI in games on Steam. Instead, it says that any action it has taken was in line with current copyright laws and policies and not an imposition of its own views. "We are continuing to learn about AI, the ways it can be used in game development, and how to factor it into our process for reviewing games submitted for distribution on Steam," the statement read. Valve further stressed that its priority has always been to ship as many of the titles it receives.
However, Valve acknowledged that the introduction of AI sometimes makes it challenging to ascertain whether a developer has sufficient rights to use AI for creating assets, including images, text, and music. It pointed out the legal uncertainty around the data used to train AI models and declared it the developers' responsibility to have the appropriate rights to launch their game. Valve maintained that while it recognizes the constant evolution of AI technology, its review process reflects current copyright laws and policies, not its own opinion.
Valve further stated that it welcomes and encourages innovation and believes AI technology will usher in new and exciting experiences in gaming. It clarified that while developers could use AI technologies in their work with appropriate commercial licenses, they must not infringe on existing copyrights. This statement comes at a time when the use of AI in creation remains a grey area globally, with governments, artists, and companies still debating the best way forward.
In a related development, Japan recently ruled that using datasets to train AI models does not violate copyright laws, as reported by Decrypt. This ruling implies that model trainers can use publicly available data without necessarily obtaining permission from the data owners. As laws and policies concerning AI and copyright continue to evolve, it remains to be seen how Valve, and indeed the wider gaming industry adapts to these developments.