Creators selling assets on itch.io will now have to be much more upfront about using generative ai. The marketplace for independent digital creators has introduced a <a target="_blank" data-i13n="cpos:1;pos:1" href="https://itch.io/t/4309690/generative-ai-disclosure-tagging” rel=”nofollow noopener” target=”_blank” data-ylk=”slk:new rule that requires users;cpos:1;pos:1;elm:context_link;itc:0;sec:content-canvas” class=”link “>new rule requiring users tag your projects if they were produced using generative ai tools, such as ChatGPT and Midjourney. Users will see an ai generation information box when they upload their projects. If they confirm that their project contains ai-generated output, they will be asked to indicate what types of content were created with generative ai, whether graphics, sounds, text and dialogue, or code.
If you have a public asset page, you'll see a dialog box when you access your dashboard, making it easier to bulk tag your projects. They will be able to select several projects from their list and then indicate whether they contain ai-generated content or not. All assets with ai output will get the “ai Generated” label, while those without will be labeled “No ai.” Each content type will also have its own subtag.
itch.io requires that all assets that used ai in some way, even if the creator edited them by hand, be labeled as ai-generated. And if you find any untagged work that used an ai tool, it will make that asset ineligible for indexing, so potential buyers can no longer find it. However, it is unclear what measures the marketplace is taking to monitor its website. While itch.io's new policy may not be enough for those who prefer to ban ai content entirely, the labels will allow buyers who don't want ai assets in their work to filter them out.
