Godot Engine Bans AI-Generated Code: Inside the June 30 Policy and the 'Slop' Crisis That Forced It

On June 30, 2026, the Godot Foundation published a blog post titled "Changes to our Contribution Policies" that formally amended the project's contribution guidelines with a dramatically stricter policy on AI-generated contributions . The policy marked the end of a period of tolerated limited AI assistance and the beginning of a hardline stance that is reverberating across the open-source and game development worlds.

The New Policy: What Changed on June 30

The core of the announcement is a clear set of prohibitions :

• No AI-authored code: Fully AI-generated code is banned from pull requests.
• No autonomous AI agents or "vibe coding": Pull requests submitted entirely or mostly by AI agents are prohibited and already trigger automated rejection.
• No AI-generated text in human-to-human communication: AI-written messages in issue comments, reviews, and discussions are also disallowed.

The Godot Foundation stated its position bluntly: "AI cannot take responsibility for code — humans must." The policy requires that every contributor be the actual human author of their submission and that any limited AI assistance used must be fully disclosed .

Alongside the AI restrictions, the Foundation also announced that new contributors — defined as those with 3 or fewer merged pull requests — are prohibited from submitting new features or significant refactoring without explicit maintainer permission. This was designed to ensure new contributors take time to learn the codebase and build trust by working on bug fixes and documentation first .

The Crisis That Led to the Policy: A Flood of AI 'Slop'

The June 30 announcement was not a sudden change of heart. It was the culmination of a crisis that had been building for months.

A Massive, Unsustainable Influx of Low-Quality PRs

By early 2026, Godot maintainers were publicly sounding the alarm. In February 2026, maintainer Rémi Verschelde reported that the project's GitHub was being flooded with AI-generated pull requests — often unlabeled as AI-produced — that were wasting enormous amounts of reviewer time just to triage and reject . He warned at the time: "I don't know how long we can keep it up" . The problem was that these submissions were frequently technically shallow, contained legally uncertain code, and their authors often lacked the understanding to fix or maintain what they had submitted .

From Tolerated Limited Assistance to a Hard Ban

Before the new policy, Godot tolerated a narrow scope of AI assistance — things like translation support, single-line code completion, regex generation, and find-and-replace tasks — as long as a human took full responsibility . But the sheer volume of autonomous AI agent submissions and "vibe coding" (where users prompt an LLM to generate entire features without understanding the output) overwhelmed that tolerance. The Foundation concluded they "can't trust heavy users of AI to understand their code enough to fix it" .

Demoralization of Reviewers and Destruction of the Mentor Pipeline

A key stated reason for the ban was the human cost. Maintainers reported that reviewing AI slop was not only time-consuming but deeply demoralizing. It diverted energy away from mentoring new human contributors — the traditional pipeline for growing the project's talent base . The policy explicitly notes that the influx of AI-generated code was actively harming the community's ability to train new developers through entry-level tasks and hands-on code review .

Copyright and Legal Liability Under the MIT License

Godot is distributed under the MIT license, which requires clean provenance for all contributed code. The Foundation flagged that AI-generated code — often trained on unknown copyrighted datasets — poses serious copyright compliance risks that could expose the project to legal liability .

Key Specific Rules at a Glance

Rule	Detail
Human authorship required	Every contribution must be the work of a human who understands the code and can maintain it .
Disclosure required	Any use of AI assistance (e.g., code completion) must be openly disclosed .
No autonomous agents	AI agents operating without direct human supervision are banned .
No vibe coding	Submissions where the human cannot explain or fix the AI-generated code are rejected .
GitHub ban threat	Contributors who continue to submit AI-authored code after warnings may face GitHub account bans .
New contributor restrictions	New contributors are prohibited from submitting new features or significant refactoring without explicit maintainer permission .

The Broader Industry Backlash: David Gaider's 'Virulent Plague' Comments

The Godot announcement came in the same week as a separate but aligned flashpoint.

In a late June 2026 interview with GamesRadar, David Gaider — the former narrative lead on Dragon Age and co-founder of Summerfall Studios — called generative AI a "virulent plague" in game development . His core arguments resonated directly with the concerns that drove Godot's policy:

• "It's not ready for prime time." Gaider said the technology remains unpredictable and unreliable for core development work. He accused executives of pushing it despite its immaturity: "There's just a lot of executives who really, really want it to be [ready]" .
• Destroys junior training pipelines. He warned that if generative AI replaces every entry-level task, "how will the next generation of developers learn their craft?" — a concern that directly mirrored Godot's own warnings about the mentor pipeline .
• Copyright and legal risks. Gaider also flagged that the provenance of AI training data creates massive legal exposure for studios that use generative AI tools .

Gaider's interview went viral across the gaming press on June 29–30, 2026, with outlets worldwide covering his comments — exactly as Godot was publishing its own hardline policy .

Summary

The Godot Foundation's June 30 policy is a direct response to a months-long crisis: a flood of unmaintainable, legally risky AI-generated pull requests that were burning out volunteer reviewers and destroying the project's ability to mentor new human contributors. After trying limited tolerance, the project shifted to a hard ban on AI-authored code, autonomous AI agents, and vibe coding — requiring full human authorship, mandatory disclosure of any AI assistance, and threatening GitHub bans for violators. The move coincided with and was echoed by David Gaider calling generative AI a "virulent plague" and warning that it is destroying the pathways for training the next generation of developers .