QGIS Bans AI-Assisted Contributions: A Hard Stance on Open Source Integrity

QGIS Takes a Hard Line: AI Contributions Officially Banned

The QGIS project, one of the most widely used open-source geographic information systems, has made a definitive move in the ongoing debate over AI in software development. On October 7, 2026, the project formally adopted QEP 428, a strict policy that prohibits AI-assisted contributions in all forms. This marks a significant reversal from its previous stance, which allowed AI assistance as long as a human remained in the loop.

The decision, merged by core developer Nyall Dawson, is a direct response to what maintainers describe as an unsustainable flood of AI-generated pull requests. The policy is not just a recommendation but a formal rule that applies to code, documentation, bug reports, comments, and all written interactions across the project's GitHub repositories.

The Failure of the 'Human in the Loop' Approach

QGIS's previous policy, QEP 408, required contributors to tag their AI-assisted work and maintain human oversight. However, the system proved ineffective in practice. According to the discussion thread, approximately 50 pull requests were tagged as 'AI Assisted,' and about 40 of them were closed without being merged. Many other AI-generated contributions likely slipped through without any tags.

Core maintainer Haubourg highlighted the core problem: "AI use from external contributors puts a massive and unbearable pressure on the shoulders of our maintainers, who already had a tough job before the era of LLMs." The sheer volume of AI-generated code overwhelmed the review process, leading to either un-reviewed pull requests or code that wasn't fully vetted for quality and security.

Why This Matters: Protecting the Human Element

Beyond the immediate logistical challenges, the QGIS team is making a philosophical stand. "Free software communities are built on human networks," Haubourg added. "QGIS is trusted by organizations because they know they can contribute back, they can hire contributors for support, training, and features. As a community project, this is something we really want to protect."

This perspective resonates with contributors like Amalvir, who admitted to submitting AI-assisted bug fixes. "I didn't realize the harm this could cause to QGIS's already fragile reviewer community," they wrote. "LLMs give you the illusion that it's possible to fix bugs in a project as complex as QGIS, but by doing this, we've broken a trust contract between the reviewer and the contributor."

The Policy's Scope and Ambiguities

While the policy is clear in its intent, its exact boundaries have sparked considerable debate. The original text uses broad terms like 'assisted' and 'used,' which some contributors worry could be interpreted too strictly. For instance, does the ban include using an LLM for brainstorming, getting code review suggestions, or even using machine-learning-based autocomplete in IDEs?

Contributor 02JanDal raised these concerns: "Using Claude Code or GitHub Copilot CLI? That's clear. But what about LLM-backed autocomplete? What about using ChatGPT to brainstorm a design? Searching for an error message and unintentionally seeing the answer in Google's AI answer?"

Tim Sutton, chair of the Project Steering Committee, offered a pragmatic interpretation: "The measure of compliance is 'did you in good conscience write the code yourself.' We are not set up to peer into your computer and look at what you did. You will know if your brain made the code you want to share or someone else's brain did."

A Deliberate 'Closed Door' Strategy

Nyall Dawson clarified that this is not intended as a permanent, unchangeable ban. "The intention is not that this becomes the 'written in stone, forever' policy," he explained. "It's rather a 'let's start with a closed door, which we can then selectively open a little if/when it's safe and proper discussion has been made on that.'"

This approach was echoed by Even Rouault, a prominent QGIS and GDAL developer, who noted, "This QEP is a stop-gap measure. Two years ago the world sucked less than today, so there's no reason to allow it without fully understanding the consequences."

Community Reaction: Mixed but Resolute

The broader community reaction, as captured by platforms like daily.dev, shows a split. About 45% of respondents have mixed feelings, 35% are skeptical, and only 20% are positive. Skeptics argue the ban is poorly defined and conflates AI-assistance with "vibe-coded slop," potentially alienating legitimate contributors who use AI as a productivity tool.

However, the project's core maintainers voted overwhelmingly in favor. The final votes came from key figures including Even Rouault, Alex Bruy, Signedav, Elpaso, and others, all casting +1 votes. The policy also received support from contributors who acknowledged the need for a moratorium to figure out a sustainable path forward.

Enforcement and Future Outlook

Enforcement relies primarily on contributor honesty, as there are no reliable technical tools to detect AI-generated code perfectly. The project has no plans to implement invasive detection mechanisms, respecting the privacy of contributors while trusting them to follow the policy's spirit.

Looking ahead, the project anticipates follow-up proposals. QEP 429, for example, suggests a "trusted contributor exception" that could create a gated allowance for experienced developers. This indicates that the ban is a starting point, not a final destination.

A Growing Trend in Open Source

QGIS is not alone in this stance. The Linux kernel warned maintainers about AI-generated patches as early as 2024. Other projects, including COSMIC desktop and OpenJDK, have also implemented restrictions on AI contributions. QGIS's explicit policy, however, sets a clear precedent for how open-source communities can respond to the challenges posed by generative AI.

For users of QGIS, nothing changes in the application itself. The policy governs how code enters the project, not how users interact with the software. Users can still write their own AI-assisted PyQGIS scripts without restriction.

The Bottom Line

QGIS's decision reflects a broader tension in software development: how to harness the power of AI without undermining the human collaboration that makes open-source projects thrive. By slamming the door shut, QGIS has bought itself time to think carefully about what it wants to let in later. As Tim Sutton noted, "Let's tackle one use case at a time so that we can all stay in the same boat and keep it rowing in the same direction."

The success of this policy will depend on whether the community can maintain trust and continue to foster meaningful human contributions. For now, QGIS has made it clear: in the world of open-source, the human element remains non-negotiable.