
What happened
System76's COSMIC desktop now requires contributors to declare: "I have not included any LLM (also known as AI) generated content in this PR, including code, comments, and descriptions." GNOME developer Michael Catanzaro argues GNOME should still accept AI-generated bug reports.
Why it matters
This sets a hard line against AI-written code in a new desktop, while a veteran developer says bot-found flaws are too valuable to ignore.
What to watch
Whether GNOME follows Catanzaro's push or keeps restrictions, which could shape AI's role across open-source projects.
WHO IT HITSOpen-source contributors to COSMIC and GNOME will face stricter disclosure rules or looser acceptance of AI-generated bug reports, potentially changing how they submit code and flaws.
Summaries like this, in your inbox every morning.
The move by System76's COSMIC desktop to ban LLM-generated contributions comes as the project nears its first stable release with Pop!_OS 24.04 in late 2025. COSMIC is a ground-up effort written in Rust, but it shares heritage with GNOME, which itself was announced in 1997 as a response to KDE's Qt licensing. GNOME remains the default desktop in Red Hat Enterprise Linux, and many GNOME developers work at Red Hat, whose management has shown enthusiasm for AI. Michael Catanzaro, a GNOME developer, has consistently pushed for accepting AI-generated bug reports, linking to earlier posts and reducing the disclosure deadline for security issues from 90 days to 30, effective August 1. The article notes that AI is finding vast numbers of bugs across large codebases, and GNOME's heavy use of memory-unsafe languages makes secure coding difficult. The debate could widen: if AI bug reports are accepted, AI-assisted triage and fixes might follow, potentially leading to AI-generated extensions and core components. This path is already controversial in KDE, but Debian and the Linux kernel have allowed AI contributions, suggesting the outcome hinges on whether GNOME's leadership aligns with Catanzaro's view.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
OpenAI published 372 mathematical results from an internal frontier model on GitHub, including improvements to…

The author built a Python pipeline using cairosvg that extracts SVG from LLM replies, statically checks format…

OpenAI released new mathematical results produced by an internal frontier model, publishing them in a GitHub r…

Google released EmbeddingGemma 2, its first natively multimodal open embedding model on the Gemma 4 architectu…

Mistral AI opened a public preview of Mistral Large 4, its 1.05 trillion-parameter MoE model nicknamed "Le Cho…

OpenAI said its internal frontier model produced 722 papers with Lean formal proofs, drawn from about 4,000 pr…
