A list of policies by different open source projects about how to engage with AI-generated contributions.
Feel free to submit Pull Requests with policies you find in other projects!
- For a similar overview with plots and more statistics, check out this excellent article by Kate Holterhoff: The Generative AI Policy Landscape in Open Source
- The CHAOSS Working Group on AI Alignment also keeps a list here: https://github.com/chaoss/wg-ai-alignment/tree/main/moderation
- A list of software that has taken steps to reject AI in its development processes: https://codeberg.org/brib/slopfree-software-index
The classification in the columns below is, in many cases, inadequate. It serves only as a general guide and should not be taken as the source of truth.
| Project | Policy link | AI/LLMs allowed? | Disclosure required? | Policy includes copyright statement? | Requires human in the loop? | Notes |
|---|---|---|---|---|---|---|
| Adwaita | Use of Generative AI | No | - | - | - | |
| Alacritty | LLM/"AI" Contributions | No | - | - | - | |
| Allium | AI Contributions | No | - | - | - | - |
| Amberol | Use of Generative AI | No | - | Yes | - | Use of AI to "answer questions and find information" permitted but strongly discouraged in favour of "existing documentation" and "chats and forums" |
| Apache Airflow | Gen-AI Assisted contributions | Yes | Yes | Yes | Yes | Add Copilot code review instructions to catch AI-slop PRs#62442 |
| Apache CouchDB | Contribution Guidelines | No | - | - | - | |
| Apache DataFusion | AI-Assisted contributions | Yes | No | No | Yes | Better ways to contribute than an AI dump |
| Apache Kvrocks | Guidelines for AI-assisted Contributions | Yes | Yes | Yes | Yes | |
| Apache PouchDB | Contribution Guidelines | No | - | - | - | |
| Apache® Software Foundation (ASF) | ASF Generative Tooling Guidance | Yes | Yes | Yes | - | Applies to all ASF projects, but enforcement is decentralized. Requires use of "Generated-by:" in commit message. |
| Arcalibre | rereading: because books are art, and art is labor | No | - | - | - | - |
| Arrow | AI-generated code | Yes | Yes | Yes | Yes | |
| Asahi Linux: | Generative AI | No | - | - | - | |
| Astropy | Usage of Generative AI | Yes | No | No | Yes | |
| attrs | Generative AI/LLM Policy | Yes | ? | Yes | Yes | "No LLM bots in Co-authored-by:s." |
| Avoyd | What is your policy on the use of generative AI and LLMs? | No | - | - | - | It is AI image denoising that uses training data with appropriate rights. It doesn't allow LLM use for writing the code. |
| Bevy | AI Policy | No | - | Yes | - | - |
| Bobby | Use of Generative AI | No | - | Yes | - | - |
| Borgmatic | AGENTS.md | No* | - | - | - | An AGENTS.md file is provided, which instructs the agent to not modify the code. AI is allowed to interact with the code in a read-only manner. |
| CapyPDF | AI Policy | No | - | - | - | |
| Cataclysm: Dark Days Ahead | CONTRIBUTING.md | No | - | - | - | |
| CCExtractor | Using AI to generate PRs before and during GSoC | Yes | Yes | Yes | Yes | Mark PRs as "AI-generated/AI-assisted/No AI used" |
| CC Open Source | Contribution Guidelines | No | - | - | - | |
| Cheese Paper | AI | No | - | Yes | - | - |
| chezmoi | Developer guide | No | - | - | - | Shows a Claude banner due to a single Claude commit that was created before the introduction of the ban. |
| Chimera Linux | Mastodon Post | No | - | - | - | - |
| ChiPass | Contributing to KeePassChi | No | - | - | - | - |
| cilium | Cilium Generative AI Policy | Yes | Yes | Yes | Yes | DCO signoff required for all contributions, including AI-generated ones. |
| Clojure | No Generated Code | No | - | - | - | |
| CloudNativePG | AI Policy | Yes | Yes | Yes | Yes | |
| conda | Generative AI policy | Yes | No | Yes | Yes | |
| CPython | Generative AI policy | Yes | No | No | Yes | "Disclosure of the use of AI tools in the PR description is appreciated, while not required." |
| CuPy | AI Policies | Yes | No | No | Yes | |
| curl | On AI Use in CURL | Yes | Yes | Yes | Yes | "The golden rule is that a contribution should be worth more to the project than the time it takes to review it, which is usually not the case if large parts of your PR were written by LLMs." |
| DataJourneyHQ | AI aware checklist | Yes | Yes | No | - | |
| Decker | Contributing | No | - | - | - | - |
| Django | AI-Assisted Contributions | Yes | Yes | No | Yes | PR template; No automated AI reviews |
| dlt | Blog post (No policy; proactively encourages AI usage) | Yes | No | No | No | |
| do | License and AI Policy | No | - | - | - | - |
| Drupal | Policy on the use of AI when contributing to Drupal | Yes | Yes | Yes | Yes | |
| Dune 3D | Use of large language models / generative "AI" | No | - | - | - | - |
| EasyBuild | EasyBuild AI Policy | Yes | Yes | No | Yes | Requires declaration of specific AI models/tools used. |
| Elastic | Use of Generative AI | No | - | - | - | - |
| Elementary OS | Generative AI Policy | No | - | - | - | - |
| EmulatorKit | ContributionRules | No | - | - | - | Explicit ban on Copilot; maintainer has confirmed this applies to other LLM contributions |
| Endless Sky | On AI-generated/assisted content | No | - | Yes | - | - |
| Evi | Using AI | No | - | - | - | - |
| FastAPI | Automated Code and AI | Yes | No | No | Yes | "If the human effort put in a PR, e.g. writing LLM prompts, is less than the effort we would need to put to review it, please don't submit the PR." |
| Fedora | AI-Assisted Contributions Policy | Yes | Yes | Yes | Yes | Fedora Council Policy Proposal: Policy on AI-Assisted Contributions |
| Fennel | Contribution Transparency | No | - | - | - | - |
| Firefox | Firefox AI Coding Policy | Yes | No | No | Yes | |
| Flutter | AI contribution guidelines | Yes | No | No | Yes | |
| Forgejo | AI Agreement | No | - | - | - | - |
| Fractal | LLM Contributions | No | - | Yes | - | - |
| gaypanel | Things that aren't allowed | No | - | - | - | - |
| GDAL | AI/LLM tool policy | Yes | Yes | Yes | Yes | |
| Gedit | Gedit Development Guidelines - No Large Language Models AI tools | No | - | - | - | - |
| Gentoo Linux | AI Policy | No | - | - | - | - |
| Ghostty | AI Usage Policy | Yes | Yes | No | Yes | "Our reason for the strict AI policy is not due to an anti-AI stance, but instead due to the number of highly unqualified people using AI." |
| GIMP | Contribution Guidelines | No | - | - | - | - |
| Gitea | AI Contribution Policy | Yes | Yes | No | Yes | |
| Gitte | Contributing | No | - | - | - | - |
| Glasgow Interface Explorer | Contributing code or documentation | No | - | - | - | - |
| GNOME Extensions | Review Guidelines | Yes | - | - | Yes | "While it is not prohibited to use AI as a learning aid or a development tool (i.e. code completions), extension developers should be able to justify and explain the code they submit, within reason." |
| GNOME Loupe | Use of Generative AI | No | - | - | - | |
| GNU Binutils | LLM Generated Content | No | - | - | - | - |
| goose | How to Use AI with goose | Yes | No | No | Yes | AGENTS.md file |
| GoToSocial | Code of Conduct | No | - | - | - | - |
| Gram | Strict No LLM / No AI Policy | No | - | - | - | Fork of the Zed editor, which has AI-generated code. However, this is explicitly a fork seeking to remove AI. The amount of AI-generated legacy code may decrease over time. |
| Halloy | AI | No | - | - | - | Has the Claude banner due to a single Claude commit. There's currently at least some effort going into removing it |
| Highscore | Use of Generative AI | No | - | Yes | - | This includes its porting interface libhighscore |
| hit-counter.net | AI Policy | No | - | - | - | - |
| Homebrew | (AI/LLM) usage | Yes | Yes | No | Yes | Disclose model/tool and only one AI-assisted PR open at a time. |
| Icechunk | AI Usage/Contribution Policy | Yes | No | No | Yes | |
| Iceshrimp | "AI" policy | No | - | - | - | - |
| Idris | Things We Will Not Accept | No | - | - | - | Has the Claude banner due to a single Claude commit that occurred before this policy took effect. After that commit was made, the project changed their contributing guidelines to block LLM code. |
| Inkscape | Mastodon Post | No | - | Yes | - | The full LLM policy is being actively developed but appears to lean towards banning generative AI |
| Inochi2D | On Generative AI | No | - | - | - | - |
| Iocaine | Contributing to the project | No | - | - | - | - |
| IREE (Intermediate Representation Execution Environment) | AI tool use | Yes | Yes | Yes | Yes | Use "Assisted-by:" or "Co-authored-by:" |
| Jellyfin | Jellyfin LLM/"AI" Development Policy | Yes | No | No | Yes | |
| Joomla | Generative AI policy | Yes | Yes | Yes | Yes | |
| Jqwik | Contributing | No | - | Yes | - | - |
| jsoup | Use of AI | No | - | - | - | - |
| Kornia | AI_POLICY.md | Yes | Yes | No | Yes | Use "AI-generated/AI-assisted/No AI" labels |
| Krita | AI Moratorium | No | - | - | - | maintainer statement, to be reviewed in October 2026 |
| Kubernetes | AI Guidance | Yes | Yes | No | Yes | "Assisted-by:", "Co-developed-by:" not allowed |
| Leiningen | Contributing | No | - | Yes* | - | "Please note that it is ethically unacceptable to submit patches (to this project or any other) which you did not author yourself without giving clear attribution to the original author. Note that this includes submitting changes generated by most so-called "artificial intelligence" language models as these systems often make it impossible to even identify (much less credit) the original author." |
| libadwaita | Use of Generative AI | No | - | Yes | - | - |
| libmanette | Use of Generative AI | No | - | - | - | |
| libxml2 | Strict No LLM / No AI Policy | No | - | - | - | - |
| Lighthouse | A quick "Constitution" for Lighthouse | No | - | - | - | - |
| LinkML | AI Covenant | Yes | Yes* | No | Yes | No co-authorship with AI tools allowed; disclose usage when you don't understand the proposed changes fully |
| Linux Kernel | AI Coding Assistants | Yes | Yes | Yes | Yes | "Assisted-by:" required; DCO required. |
| Linux man-pages | AI - artificial intelligence policy | No | - | - | - | |
| LLVM | LLVM AI Tool Use Policy | Yes | Yes | Yes | Yes | "Using AI tools to fix issues labelled as “good first issues” is forbidden." |
| Lobsters | Getting oriented | No | - | Yes | - | - |
| LÖVE | Contributing | No | - | - | - | |
| Makko | Contributor Policy Agreement | No | - | - | - | - |
| Matplotlib | Restrictions on Generative AI Usage | Yes | Yes | No | Yes | "AI Disclosure" secion in PR template must be filled |
| MDAnalysis | MDAnalysis AI-generated contributions policy | Yes* | Yes | No | Yes | No "substantial" contributions generated by AI tools. |
| Mesa | Submitting Patches | Yes | Yes | Yes | Yes | Use "Assisted-by:" or "Generated-by:"; do not use "Co-authored-by:" with AI tools. |
| mGBA | Contribution Guidelines | No | - | - | - | - |
| MicroPython | Generative AI Policy | Yes | No | No | Yes | |
| Migraine Log | Use of Generative AI | No | - | Yes | - | - |
| Mikan | Contributing | No | - | - | - | - |
| MirBSD | More on the so-called “AI” craze | No | - | Yes | - | - |
| Molecular Nodes | MolecularNodes AI-generated contributions policy | Yes* | Yes | No | Yes | No "substantial" contributions generated by AI tools. |
| Musl | Re: Proposed "AI" policies | No | - | Yes | - | Threats reading an AI generated solution as being potentially tainted by proprietary code. |
| napari | AI use policy and guidelines | Yes | Yes | No | Yes | |
| NetBSD | NetBSD Commit Guidelines | Yes* | Yes | No | No | AI-assisted contributions require explicit core approval |
| NumPy | AI Policy | Yes | Yes | Yes | Yes | Mailing list discussion |
| nvim-tree | AI Usage Policy: Highly Discouraged | Yes* | Yes | No | Yes | "AI generated code is discouraged as this doesn't match (these) nvim-tree values."" |
| OCaml | Guidelines relevant to AI-assisted contributions | Yes | Yes | Yes | Yes | |
| Open edX | AI Contribution Policy | Yes* | Yes | No | Yes | Only specific tools with a "sufficient reputation for proper training" are allowed. |
| OpenInfra | Policy for AI Generated Content | Yes | Yes | Yes | Yes | Required use of "Assisted-by:" or "Generated-by:". Open Source AI models recommended. |
| OpenJDK | OpenJDK Interim Policy on Generative AI | No | - | - | - | Interim policy. |
| OpenTTD | Use of AI | No | - | ? | - | "Additionally, AI-generated code conflicts with this project's license (GPL v2), since you cannot truly release code for use if you didn't author it yourself." |
| Oxide | Using LLMs at Oxide | Yes | - | No | Yes | Comprehensive, extensive policy. |
| Pallets | LLM and AI Policy | No | - | - | - | Applies to Flask, Jinja, Click, etc. "This policy focuses on trust and quality, but there are numerous other issues with the social, ethical, resource, and environmental impacts of LLM development and use." |
| Pandas | Automated contributions policy | Yes | Yes | No | Yes | |
| Pandemonium Engine | Contributions: Use of AI content generators | No | - | Yes | - | - |
| pgwatch | ARTIFICIAL INTELLIGENCE & AUTOMATION POLICY (V1.0) | Yes | Yes | No | Yes | Must mention specific tools used. |
| pip | Generative AI/LLM Policy | Yes | No | Yes | Yes | No "Co-authored-by:" with AI tools. |
| pip-tools | LLM Generated Contributions | Yes | No | No | Yes | |
| pkgconf | Instructions for agentic systems | No | - | - | - | Instructs agents to stop. |
| Poezio | Use of Generative AI | No | - | Yes | - | - |
| Polars | AI Usage Policy | Yes | Yes | No | Yes | |
| Polyphony | Polyphony Contributor Code of Conduct | No | - | Yes* | - | "The training data used for these models is largely obtained without the consent of artists, denying them their right of control over their work and any sort of financial compensation." |
| postmarketOS | AI Policy | No | - | - | - | - |
| Processing/p5.js | AI Usage Policy | Yes | Yes | No | Yes | Disclose specific tools used. |
| Pytest | AI/LLM-Assisted Contributions Policy | Yes | No | No | Yes | Recommended use of "Co-authored-by:". |
| PyTorch | AI-Assisted Development | Yes | No | No | Yes | |
| PyVista | Generative AI | Yes | No | No | Yes | Follows CPython's policy. |
| QEMU | Use of AI-generated content | No | - | - | - | - |
| QGIS | QGIS Enhancement: AI tool use policy | Yes | Yes | Yes | Yes | Use "Assisted-by:" or "Generated-by:" labels. |
| qutip | AI Tools Usage Policy | Yes | Yes | Yes | Yes | No AI contributions to "good first issues". |
| Redox OS | AI Policy | No | - | - | - | - |
| Requests | Generative AI/LLM Policy | Yes | No | Yes | Yes | No "Co-authored-by:" with AI tools. |
| ruff | AI Policy | Yes | No | No | Yes | |
| Rust | LLM Usage Policy | Yes | Yes | No | Yes | Announcement (blog post) |
| Schism Tracker | AGENTS.md | No | - | Yes | - | "Code generated by such LLMs may be based upon code under a license that is incompatible with the GPLv2." |
| SciActive | Human Contribution Policy | No | - | - | - | - |
| scikit-image | AI Policy | Yes | Yes | No | Yes | |
| scikit-learn | Automated Contributions Policy | Yes | Yes | No | Yes | |
| SciPy | AI Policy | Yes | Yes | Yes | Yes | |
| SDL | PR Template | No | - | - | - | AGENTS.md file |
| SDL_net | AGENTS.md | No | - | - | - | Same policy as SDL. |
| SearXNG | AI Policy | Yes | Yes | No | Yes | |
| Sentinel | AI guidelines | No | - | - | - | |
| Servo | AI Contributions | No | - | - | - | |
| Skarnet | Re: [PATCH] Add CLAUDE.md | No | - | - | - | - |
| Smederee | Contribution guide | No | - | - | - | Point in section "Patch Requirements": "A patch SHALL NOT include any contributions which are believed to include or derive from AI generated content." |
| snac | Contributing to snac | No | - | - | - | - |
| Sphinx | AI Policy | Yes | Yes | Yes | Yes | - |
| STAC | AI/LLM tool policy | Yes | Yes | No | Yes | Use "Assisted-by:" and similar labels. |
| stb | AI AND LLM ARE FORBIDDEN | No | - | - | - | |
| SymPy | AI Generated Code and Communication Policy | Yes | Yes | Yes | Yes | |
| Teamtype | Generative AI/LLM policy | No | - | - | - | - |
| Telegraf | InfluxData Policy on AI-Generated Code Contributions | Yes | No* | Yes | Yes | Disclosure only required upon being asked. |
| Tenacity | The Use of LLMs and AI in Contributions | No | - | Yes | - | - |
| TorchGeo | AI Policy | Yes | Yes | Yes | Yes | - |
| Tusky | AI policy | No | - | - | - | - |
| ty | AI Policy | Yes | No | No | Yes | |
| typescript-eslint | AI Contribution Policy | Yes | No | No | Yes | |
| Unbound | LLM Policy | No | - | - | - | This policy covers all NLnet Labs projects such as ldns, NSD and Rotonda. LLM use permitted to generate vulnerability reports, to translate when communicating and for local assistance only. |
| Unison | LLM usage | No | - | - | - | - |
| University of Alaska Anchorage | GSoC/Acceptable-and-Ethical-AI-Use-Policy.md | Yes* | No | No | Yes | "No vibe code" |
| uv | AI Policy | Yes | No | No | Yes | |
| Veloren | Contributing | No | - | Yes | - | "Content generated by large language models [...], is presumed to be tainted content, and must not be committed." |
| Vim Classic | Free of generative AI | No | - | - | - | - |
| Wafrn | AI Contributions Policy | No | - | - | - | - |
| Wagtail | Use of generative AI | Yes | Yes | No | Yes | |
| Wikipedia | Artificial Intelligence | No* | - | - | - | Allows a few, very specific uses such as translation assistance. |
| WP1 | Usage of LLMs/AI coding assistants | Yes | No | No | Yes | |
| x86css | What ai/model/llm do you use? | No | - | - | - | "I do not use ai. I don't think you can build a project like this with an llm." |
| Xarray | AI Usage Policy | Yes | No | No | Yes | Xarray's CLAUDE.md file |
| XScreenSaver | FAQ: 25. | No | - | - | - | - |
| yt-dlp | AUTOMATED CONTRIBUTIONS (AI / LLM) POLICY | No | - | Yes | - | "AI-generated code conflicts with this project's license (Unlicense), since you cannot truly release code into the public domain if you didn't author it yourself." |
| Zig | Strict No LLM / No AI Policy | No | - | - | - | Contributor Poker and Zig's AI Ban - rationale for ban |
| Ziglings | AI-Generated Content | No | - | - | - | AI tools are allowed to polish English only |
| Zulip | AI use policy and guidelines | Yes | No | Yes | Yes |
- Agda: Blanket ban on LLM-generated contributions#8456
- Blender: AI Contibutions Policy
- Debian: General resolution: LLM usage in Debian: First call for votes (Summary of discussions on lwn.net)
- FreeBSD: Policy on generative AI created code and documentation
- GCC: Working Group for GCC AI Policy
- Inkscape - Policy Discussion: Generative AI
- Jupyter: AI-assisted code policy: "Literature Review"
- musl-libc: Tentative policy
- NixOS/nixpkgs: How do we deal with AI-generated issues? #410741
- Perl: An LLM policy
- Linux Foundation's Guidance Regarding Use of Generative AI Tools for Open Source Software Development
- Preparing JOSS for a generative AI future: From code to human creativity and design
- Navigating LLMs in Open Source: pyOpenSci's New Peer Review Policy
- scikit-learn's AGENTS.md file
- agentic-oss-policy: Policy templates for protecting open source projects from autonomous AI agent abuse
- llama.cpp AGENTS.md file
- Pi (AI agent toolkit) – AGENTS.md file
- The "Vouch" system for managing trust in contributions
- Good Egg: Trust scoring for GitHub PR authors
- Processing AGENTS.md file
- p5.js AGENTS.md file
- Position Statement on Generative AI in the RSE Workplace
- Exploring Solutions to Tackle Low-Quality Contributions on GitHub
- LibreHealth GSoC participation policy: README: The use of AI is forbidden at any point during the program
- Superpowers CLAUDE.md file
- rOpenSci's policy on the use of AI on software peer review
- Linux Foundation Report: Open Source and the Future of AI
- Responsible AI Guide
- Let's talk about AI slop
- Scientific Python: Blog post: Community Considerations Around AI Contributions
Many thanks to multiple contributors who shared links and contributed in private, on Mastodon or LinkedIn (let me know if you want attribution 😄)