Flathub just reversed its AI-generated content ban, but developers still need to say they used it

Published Oct 8, 2026, 7:14 PM EDT Simon is a Computer Science BSc graduate who has been writing about technology since 2014, and using Windows machines since 3.1. After working for an indie game studio and acting as the family's go-to technician for all computer issues, he found his passion for writing and decided to use his skill set to write about all things tech. Since beginning his writing career, he has written for many different publications such as WorldStart, Listverse, and MakeTechEasier. However, after finding his home at MakeUseOf in February 2019, he would eventually move on to its sister site, XDA, to bring the latest and greatest in Windows, Linux, and DIY electronics. Sign in to your XDA account Summary Flathub now allows AI-generated code if disclosed and labeled with the affected parts and extent. Submitters must be human; agents can't generate commits, descriptions, reviews, or replies. Maintainers can still reject AI-labeled or fully vibe-coded apps if they are low quality or problematic. Back in May of this year, we reported that Flathub banned AI-generated submissions because people were being a little rude about it. Now, as more and more Linux distros have declared acceptance of LLM-created code (as long as it's of high quality, of course), it seems that Flathub has had a change of heart. Fortunately, while the new rules do allow AI content if it's properly disclosed, it doesn't open the door for people to flood the repository with poorly-made apps. Flathub now allows AI-generated apps, but they still have to be good People can't get out of submitting bad apps just because they said an AI made it As spotted by Linuxiac, Flathub has gone back on its ban on LLM-created content. Now, in a revised commit to the documentation, the rules begin as follows: Submitters must disclose any AI-generated code, documentation, packaging, or other material they know or reasonably believe is included in the application or its Flathub packaging. The disclosure must identify the affected parts and approximate extent. This does not mean someone can tell an LLM to make an app in five minutes, slap a disclosure on it, submit it, and earn immunity from being taken off Flathub. If the maintainers believe a project was entirely vibe-coded, or they have "concerns about its review, quality, or maintainability," they can reject an app even if the submitter correctly labelled their work. It seems to be a push to encourage usage of AI tools without opening the door to a tidal wave of vibe-coded apps that overwhelm maintainers. On top of that, people cannot just spin up an AI agent and have it maintain their project for them. Humans must submit the work themselves, and an agent cannot "generate their commit messages, descriptions, review comments, or replies." This is likely to stop people from parking an agent on a project and having it do all the maintenance work.