QGIS bans AI-assisted code contributions
“No LLM, AI or agent assisted code contributions are allowed”
Confirmed
The takeaway
The open-source geographic information system QGIS replaced its ‘human in the loop’ AI policy with an outright ban on AI-assisted code, with code reversion and loss of contributor rights as sanctions. Its lead proposer said the earlier policy had failed in practice.
Status
- Claim
Confirmed
- Our reporting
- High confidence
- Importance
- 2 of 5
- Last verified
- 12 October 2026
Your AI and this story
- GPT-6 Astra160 days after its cutoff
- Claude Opus 5.599 days after its cutoff
- Gemini 3.8 Flash190 days after its cutoff
- Grok 4.7129 days after its cutoff
None of these four assistants can know about it. The closest, Claude Opus 5.5, stops 99 days before it.
Key facts
- Pull request: ‘Add QEP428: No ai contributions permitted’, qgis/QGIS-Enhancement-Proposals #387, opened 24 Sept 2026 by nyalldawson, merged 7 Oct 2026 (GitHub API)
- PR description: ‘This formally changes the AI policy to a strict “NOT ALLOWED” policy.’
- New policy text (file qep-408-ai-tool-policy.md, version 2.0, author ‘the QGIS project contributors’): ‘No LLM, AI or agent assisted code contributions are allowed in the https://github.com/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.’
- Removed policy (by Even Rouault): ‘contributors can use whatever tools they would like to craft their contributions, but there must be a human in the loop’, with labelling of tool-generated content and a ban on agents acting without human approval
- Discussion on the PR (as summarised from the page, not re-read line by line): Dawson said the human-in-the-loop approach had failed in practice and proposed to ‘start with a closed door, which we can then selectively open a little if/when it’s safe’; others asked whether autocomplete, translation, review and security auditing fall under the ban
- Scope stated in the text: code contributions to the main qgis/qgis repository
- Reach: Hacker News front page on 12 Oct 2026, 344 points
What happened
QGIS is a widely used open-source desktop geographic information system. Its earlier policy (QEP 408, written by Even Rouault) allowed AI tools if a human reviewed the output, labelled it and could answer questions about it. That document’s warning said it was “not currently a settled question within the QGIS community whether code created by LLMs is acceptable in the QGIS codebase at all”.
After a mailing-list discussion in September 2026, Nyall Dawson proposed replacing that policy with a ban. The change was voted on and merged on 7 Oct. The new text is three sentences long and adds sanctions: code found to be AI-assisted can be reverted and the contributor’s rights restricted.
Why it matters
QGIS first tried a disclosure-and-accountability rule and then replaced it with a ban, which makes it a test case for other projects weighing the same choice. The stated reason is maintainer load and trust between reviewers and contributors, not code quality alone. The ban also raises an enforcement question the discussion did not settle: where assistance (autocomplete, translation, review) ends and AI authorship begins.
Sources
3 sources from 3 sites. Numbers match the chips in the text.
3 sources: 1 primary, 2 reactions
Primary
- QGIS Enhancement Proposals PR #387: Add QEP428: No ai contributions permittedgithub.com, official
Reactions
- qgis-developer mailing list thread (Sept 2026)lists.osgeo.org, discussion
- Hacker News discussionnews.ycombinator.com, discussion
Changes
- Filed from the pull request (GitHub API: title, dates, diff) and the PR discussion