Flathub Reverses AI-Generated App Ban

Flathub has backed off the ban it introduced in May on applications containing AI-generated or AI-assisted material. Such applications can now be submitted, provided the developer discloses what was generated and how much of it there is.

The New Rules

The updated requirements, drafted in flathub-infra/documentation#641, read: "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."

What happens after disclosure is left to the reviewers. They can reject an application over the amount or role of AI-generated content, over code quality, or because reviewing and maintaining it would be too difficult, and a submission can be turned away without further review. The requirements only cover material that ends up in the application or its packaging; AI used for research, discussion, or debugging during development needs no disclosure.

What Remains Forbidden

Manifests are still off-limits. The policy states that Flathub manifests "must not contain AI-generated or AI-assisted content" and that disclosure does not exempt them from this restriction. The same goes for the submission process itself: AI tools may not open or automate submission pull requests, generate commit messages and descriptions, or produce review comments and replies, and requesting AI-agent reviews is forbidden. Submitters must be human.

Developers who fail to disclose AI-generated material, misrepresent its extent, or otherwise violate the submission rules risk rejection. Repeated violations can lead to a permanent ban from future submissions.

The May policy had rejected AI-assisted applications outright, with possible exemptions only for mature, well-maintained projects. How many disclosed submissions actually get through under the new rules is something Flathub has not published numbers on yet.