The GNOME Shell Extensions store has updated its review guidelines to explicitly ban AI-generated extensions, marking a significant step in the open-source community’s approach to artificial intelligence in software development. The change, announced earlier this month, states that extensions must not be primarily created using AI, though developers are still permitted to use AI tools as assistants during the development process. Javad Rahmatzadeh, a developer who reviews extensions for GNOME, explained in a blog post that the policy shift was prompted by a surge in submissions containing low-quality, poorly structured code generated by AI. He noted that many of these packages included excessive, redundant lines of code and violated established best practices, making them difficult to maintain and potentially harmful to system stability. GNOME extensions serve a wide range of functions, from minor visual tweaks to essential system enhancements. One of the most widely used, Dash to Dock, offers users greater control over the application dock, demonstrating how critical well-crafted extensions can be to the overall user experience. The new guidelines emphasize that while AI can be used as a learning aid or to help with code suggestions, developers must fully understand and manually review any AI-generated code before submission. The goal is to ensure that extensions are reliable, secure, and maintainable—qualities essential for a stable desktop environment. This move aligns with broader concerns in the open-source world about the quality and integrity of code produced with AI assistance. Earlier this year, the Fedora Council introduced similar rules around "vibe coding," though it did not implement a full ban. GNOME’s stricter stance reflects a growing awareness of the risks associated with unchecked AI-generated software, especially in systems where reliability and security are paramount.