QGIS Maintainers Ban AI Contributions to Reduce Review Burden
The open-source mapping software project is enforcing a strict no-AI policy to protect maintainers from the overwhelming volume of low-quality pull requests.
Key points
- QGIS has formally updated its policy to strictly prohibit AI-generated contributions.
- Maintainers cite an unsustainable review workload caused by the high volume of LLM-assisted code.
- The ban aims to preserve the human-centric trust network that organizations rely on for support.
The QGIS project has formally adopted a strict policy prohibiting any contributions generated by artificial intelligence tools. According to a pull request on the QGIS-Enhancement-Proposals repository, the project is moving beyond its previous "human in the loop" stance to a complete ban on AI-assisted submissions.
This policy change addresses the massive pressure external AI use places on project maintainers. Developers note that the volume of AI-generated code exceeds the capacity for serious human review. This imbalance leads to either long backlogs of unreviewed pull requests or insufficiently reviewed code that introduces quality issues.
Rationale for the Policy Shift
The decision stems from a concern that AI tools create an illusion of competence for contributors who may lack deep understanding of the codebase. One contributor noted that using large language models to fix bugs broke the implicit trust contract between reviewer and contributor. The maintainers argue that this trust is essential for a project of QGIS's complexity.
QGIS is trusted by organizations because they know they can contribute back to the project and hire contributors for support and training. The community views this human network as a core asset. Protecting the bandwidth of maintainers is seen as necessary to keep the project welcoming to genuine human contributions.
Impact on Contributors and Reviewers
The new policy applies to all external contributions, aiming to reduce the strain on the already fragile reviewer community. Contributors who previously used AI to investigate bugs or suggest fixes are now required to submit work that reflects their own understanding. This shift acknowledges that while AI can produce fixes quickly, it does not ensure the contributor knows what they are doing.
Discussions regarding the use of LLMs for review, auditing, and security purposes are being treated as separate topics. The project team plans to open further discussions about these uses at the upcoming user conference. For now, the focus remains on enforcing the strict ban on AI-generated code submissions.
What should you do?
Contributors to QGIS should review the updated policy before submitting any pull requests or issue comments. Ensure that all code and documentation you submit is written by you and reflects your own technical understanding. Watch for further announcements regarding the use of AI in review and auditing processes at the user conference.
Background: Large language models
Large language models are a specific type of neural network trained on text. Generative AI is the broader umbrella term for any system that creates new content, including images and code. Not all generative AI uses large language models, but most modern text generators do.
What to do and how to stay safe: QGIS
- Review the QGIS contribution guidelines to ensure compliance with the new no-AI policy.
- Verify that all submitted code and documentation is authored entirely by human contributors.
- Monitor official QGIS communication channels for updates on AI usage in review processes.
- Engage in community discussions to understand the rationale behind maintaining a human-centric network.
Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
Does the ban cover all uses of AI in the QGIS project?
The current ban covers contributions from external contributors. Separate discussions on AI for review and auditing are planned for the user conference.
Why is QGIS implementing such a strict policy?
The project aims to reduce the unsustainable review burden on maintainers and preserve the quality and trust of human contributions.
When will this policy take effect?
The policy change is formalized through the QGIS-Enhancement-Proposals pull request, indicating it is now in effect for new submissions.
