Debian Developers Endorse Responsible AI Use

Published the results of the general vote (GR, general resolution) of Debian project developers involved in package maintenance and infrastructure maintenance, which addressed the issue of using large language models and AI tools in the development of the distribution. The fifth point “Responsible use of generative AI” won the vote.”

The selected option does not endorse, but does not prohibit the use of AI. It is proposed to apply the same quality standards and licensing requirements to code created through AI as to code written manually. It is not allowed to transfer the results of AI work without checking, testing and understanding the essence of the change. All responsibility for changes prepared through AI falls on the developers who transferred them to the project.
Copyright and licensing issues for works created through AI are addressed under existing Debian policies.

Disclosure of information about the use of AI is encouraged, but not required. It is prohibited to use third-party AI services to work with confidential data, such as access keys and credentials, as well as to analyze vulnerabilities that have not yet been publicly disclosed. Carrying out mass automated actions using AI requires prior discussion and must be performed under human supervision.

/Reports, release notes, official announcements.