Technology media outlet Linuxiac reported today that the Debian community recently voted on a proposal concerning large language models. The results showed that most developers supported the responsible use of generative AI.
The vote included nine options, ranging from a complete ban on large-language-model-assisted development to allowing AI use under specific conditions. The result will affect how Debian developers work and how development teams respond to new AI tools in the future.
The vote used the Condorcet method, with option 5, “Responsible use of generative AI,” ultimately winning.
At the same time, Debian’s new policy does not prohibit using generative AI in software development or documentation workflows. The policy recognizes that responsible use of AI can improve productivity and allow developers to focus more on other tasks.
However, the new policy requires all AI-assisted code and documentation to meet quality and legal compliance requirements, with human developers bearing ultimate responsibility.
In addition, Debian will not require contributors to disclose whether their code was created using AI. Although the project encourages disclosure, it does not make it mandatory.