- Version: 1.1 (2026-02-17)
- License: CC-BY-4.0
This document defines the MDAnalysis organization's policy regarding AI-generated content. This policy applies to all aspects of the MDAnalysis project, including all the GitHub repositories under the organization.
For the purposes of this document, content is defined as any contribution to the MDAnalysis organization, both code and non-code, including but not limited to:
- Code submitted for inclusion in MDAnalysis, or other organization-owned repositories
- Documentation of any kind
- Communication content (e.g. issues, discussions, pull request descriptions, and reviews)
We recognize that the use of AI in open source software raises many difficult questions and acknowledge that views, approaches, and technical capabilities are rapidly changing. We invite comments and further discussion (via our discussion channels) and expect to periodically review this policy.
MDAnalysis does not accept any substantial uses of AI-generated content in contributions.
AI tools may be used in limited cases, for example when it is used for small amounts of AI-assisted code (e.g. autocompletion using copilot, cursor, etc...) or for language translation purposes. Such uses must always be declared.
Please understand the reviewing capacity is very limited for the project, so large PRs which appear to not have the requisite understanding might not get reviewed, and eventually closed or redirected. Understanding from authors about what code they are contributing and how it fits into the larger picture is crucial. If reviewers feel there is a lack of understanding from looking at the code, PR descriptions or just general interactions with a contributor, we reserve the right to close the PR to save time and resources. Given the limited reviewing capacity, time and effort of reviewing and educating contributors will be more targeted to PRs of initial high quality and/or contributors where interactions are constructive and of good quality.
All contributions must be fully transparent about any AI usage. Failure to disclose AI use (e.g. in pull request description details) may be queried by other developers and may be considered non-compliance with this policy.
If we suspect that AI tools were used as part of a contribution, MDAnalysis developers may ask you for clarification. Please do not take offense if this happens to you, distinguishing between AI-generated and human-authored content is increasingly difficult!
Fully AI-generated contributions, such as content generated by prompting a large language model (e.g. claude code or chatgpt) or produced by any agentic AI system are not permitted in MDAnalysis.
Some limited exceptions are described in sections #2 and #3 for integrated development environments (IDEs) and language translation.
AI tools are now so deeply integrated into many modern development environments, and it may be impractical to avoid them entirely. We recognise that this is particularly true for modern integrated development environments (IDEs), which may use AI to power features like autocompletion.
In these limited cases of code generation, typically ranging from a single variable to a couple of lines of code, AI assitance is deemed acceptable. However, if code generation exceeds minimal, sporadic amounts (e.g. repeated or large multi-line blocks), it would be considered fully AI-generated and, as defined in section #1, is not acceptable.
As per section #0, it is your responsibility to report the extent of AI assitance you used.
MDAnalysis aims to foster an inclusive environment for all contributors and users. We recognise that language barrier can limit participation in open-source projects, particularly for non-native English speakers.
Whilst we prefer you communicate with us in your own words (and languages if needed), we acknowledge that AI-based translation tools may sometimes be necessary for communicating in issues / discussions. If such tools are used, we ask you to disclose their use and purpose.
Contributors are responsible for the content and provenance of every aspect of their contribution, including any design decisions. Contributors must be aware of whether their work derives from external sources / codebases and must declare any derivative work.
All code merged into MDAnalysis repositories must be reviewed by a human reviewer. Instructions / suggestions from human reviewers always take precedence over those of non-human reviewers.
Any failures to comply with this policy should be reported to the MDAnalysis core developers through our standard communications channels or email to mdanalysis@numfocus.org . The core developers may enforce this policy through a range of actions, which may include restricting or banning individuals from future contributions.
There is a zero tolerance policy for intentionally failing to disclose or misrepresenting the extent of AI usage.
We acknowledge the Kornia AI policy and discussions within the Scientific Python Community for providing guidance and insight in shaping this policy.