The real story behind the Godot open source AI ban
The Godot open source AI ban is a policy decision by the Godot Foundation to prohibit most AI-authored contributions because they erode code quality, maintainability, and the human mentoring relationships that keep volunteer projects alive.
The Godot Foundation is rewriting its contribution guidelines to prohibit AI-authored code, pull requests submitted by autonomous AI agents, and AI-generated text in human-to-human communication. This move follows months of internal discussion as maintainers face a growing backlog of pull requests, many of them produced by code-generation tools. AI submissions have overwhelmed a pipeline already constrained by a small number of qualified reviewers and the demanding nature of code review work. The new rules still allow limited AI help for menial tasks like code completion, regex, or find-and-replace, as long as contributors disclose it. In short, Godot is not rejecting AI outright; it is drawing a hard line between AI as a spell-checker and AI as a co-author, and that distinction is a clear statement about what kind of community it wants to be.
Demoralizing slop: what Godot is really banning
On paper, this looks like a policy about AI-generated code quality. In practice, it is a revolt against what maintainers describe as low-effort slop backed by no accountability. Godot’s maintainers say AI contributions have flooded the review queue faster than they can keep up, turning an already heavy workload into something unmanageable. The Foundation explicitly calls AI contributions “demoralizing,” because feedback on these pull requests does not change how a model behaves next time, and it cannot trust heavy AI users to understand their own code well enough to act on that feedback. The technical complaints are familiar: highly variable AI-generated code quality, gaps in accountability, and contributors who cannot fix what they submitted once it breaks. What Godot is banning is not experimentation with tools; it is the flood of anonymous, responsibility-free code that treats maintainers like unpaid QA for a model.
How AI submissions undermine open source mentoring
The more important story is not about syntax errors; it is about mentorship. Godot’s maintainers describe reviewing pull requests as tedious but rewarding because their effort goes into educating a contributor who might become a future maintainer. That reward disappears when the “contributor” is a model. AI contributions break this delicate loop: detailed feedback ends up feeding a machine that cannot learn, cannot grow, and will never step up to maintain the project. One maintainer from another project called this dynamic “contributor poker”: in open source, you bet on the person, not the contents of their first pull request, and an AI-generated pull request breaks that calculation entirely because there is no human on the other end to grow from review. In effect, AI submissions turn mentorship time into free tuning data, while starving the community’s pipeline of future human maintainers.
New rules for contributors: raising the bar on responsibility
Godot’s open source AI ban comes with a second, quieter change that will shape developer workflows: new contributors—defined as anyone with three or fewer merged pull requests—are now prohibited from submitting new features or significant refactors without explicit maintainer permission. This is a blunt response to backlog pressure and a statement about developer responsibility. If you want to shape the engine, you first earn trust through bug fixes and documentation. The Foundation still allows AI for “menial things” like code completion and regex, but any use must be disclosed, and substantial AI-generated code is outright banned. AI-generated text in human-to-human discussions is also off-limits, with an exception for machine translation of human-authored content. This is not tool-phobia; it is an insistence that someone clearly owns every line that lands in the repository and can be held responsible when it misbehaves.
What this means for developers and the future of AI-assisted coding
Godot is not alone. Other open source projects have tightened or closed their pipelines over similar concerns, from AI-generated slop to false bug reports and mentorship collapse. These moves signal a broader tension: AI-assisted development promises speed, but open source relies on slow, human trust-building. AI agents consume reviews, feedback, and mentorship without participating in the social contract that keeps volunteers engaged, and ignoring that dynamic would erode the system from the inside. For developers, the message is clear: AI is welcome as a clerk, not as a ghost author. You remain responsible for understanding, defending, and maintaining what you submit. The Godot Foundation calls its current approach “conservative” and expects to revisit it as tools change, but the principle is unlikely to move: models cannot take responsibility, so projects will keep insisting that humans do.
