QGIS, one of the world's most widely used open-source geographic information system applications, has formally adopted a strict ban on AI-assisted code contributions — a policy that went into effect when its governing proposal was merged on October 7 and is now drawing renewed attention after a lively Hacker News discussion on Monday.

The change, tracked as QEP 428 ("No AI contributions permitted") and merged by longtime QGIS core developer Nyall Dawson, rewrites the project's earlier AI tooling policy from a human-in-the-loop framework into an outright prohibition. It follows a debate on the project's qgis-developer mailing list in September.

What the Policy Says Now

The updated policy document, version 2.0, states plainly:

"No LLM, AI or agent assisted code contributions are allowed in the qgis/qgis repository. If we subsequently discover that you have used an AI, LLM, or any agent-based workflow for your contribution, you will be subject to code reversion and contributor rights sanctioning."

That is a considerably harder line than the policy it replaced. The previous version, QEP 408, published in January 2026 by Even Rouault, had permitted AI use under strict conditions: contributors had to review all generated content themselves, disclose substantial tool-generated content using an "Assisted-by" commit trailer, and be able to answer questions about their work during review. It also banned autonomous agents — such as GitHub's @claude bot — from taking actions in the project's spaces without human approval.

Under the new policy, that entire framework is gone. There is no disclosure mechanism, no permitted use case, no human-in-the-loop accommodation. The policy file's authorship line now simply reads "the QGIS project contributors."

Why a Mapping Project Drew a Line

QGIS is a cornerstone of the open-source geospatial stack, offering professional-grade desktop GIS capabilities used by governments, environmental agencies and businesses worldwide. Like most community-driven open-source projects, it depends on a small pool of volunteer core contributors who review incoming code.

That reviewer burden is central to the ban's rationale. The project's earlier policy had already noted that passing maintainer feedback to an LLM "doesn't help anyone grow, and does not sustain our community." The September mailing-list discussion and the pull request's 25 comments show maintainers weighing whether disclosed, human-reviewed AI code still imposed unsustainable costs on scarce reviewer time — and concluding that it did.

The move taps into a broader anxiety running through open-source communities: that machine-generated contributions, however fluent, shift effort from writing code onto human reviewers who must verify it, and that disclosure-based policies depend on honesty the system cannot enforce. QGIS's answer is enforcement by consequence — reversion of merged code and sanctions against contributor rights.

A Signal for the Open-Source Ecosystem

QGIS is not the first project to restrict AI contributions, but the totality of its ban — covering code, and by extension the agent-based workflows increasingly built into development tools — makes it one of the clearest statements any major open-source project has made on the question. The Hacker News thread, which reached nearly 200 points within hours on Monday, quickly filled with debate over whether such bans are sustainable as AI assistance becomes embedded in every editor and terminal.

The policy also highlights a widening split in the software world. Commercial projects increasingly measure and accept AI-generated code, betting on AI code review to police quality. Community-run projects, whose reviewers are volunteers and whose incentive structures differ fundamentally, are in some cases reaching the opposite conclusion: that the cost of verifying machine-written code outweighs the benefit of producing it.

For now, contributors to QGIS face an unambiguous rule: whether an AI touched the code is no longer a disclosure question — it is a compliance question, and the answer must be no.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →