The Godot Foundation, responsible for the popular open-source game engine, has announced a significant policy shift that will prohibit AI-authored code and AI-generated text in human-to-human communications from contributors. This decision comes after months of deliberation concerning a rising tide of "AI slop" pull requests that have been taxing the project's volunteer maintainers.
In a recent blog post, the Foundation stated that while an increase in contributions signals growing interest in Godot, the influx of AI-generated submissions has become increasingly draining and demoralizing for code reviewers. The core issue, according to the Foundation, is accountability. "AI cannot take responsibility, and we can’t trust heavy users of AI to understand their code enough to fix it," the statement read.
The Godot Foundation is in the process of updating its contribution policies to establish clearer guidelines and add barriers to low-effort submissions. The aim is to ensure that contributions come from human developers who can be held responsible for their code and are capable of fixing it if issues arise. This focus on human accountability is seen as crucial for mentoring new contributors and developing future maintainers.
While AI assistance for "menial things" will still be permitted, contributors will be required to disclose its use. Similarly, AI-generated text in communications between human contributors will be rejected as a matter of basic respect, though machine translations of human-authored text will remain acceptable.
The Foundation acknowledged that the landscape of AI tools is constantly evolving and pledged to maintain a conservative approach while re-evaluating its policies as technology progresses. This move by Godot highlights a growing concern within open-source communities regarding the integration and management of AI-generated content and code.