If generative AI made a substantial contribution to your code, say so—and describe the contribution plainly. That is Scott Donaldson’s argument in his September 11, 2026 LinuxLinks essay, not a universal rule for every software project. The practical standard is to follow the project’s own policy, make the extent of assistance clear, and take responsibility for reviewing everything you submit.
Why disclose substantial AI assistance?
Donaldson’s case is about provenance: people evaluating a project may want to understand how it was developed and what context could matter to its future maintenance. He distinguishes substantial generation—such as functions, tests, documentation, refactors, or larger portions of an application—from routine use of an editor or linter. The point is not that AI-written code is inherently bad, or that disclosure proves code is good. It is that substantial AI involvement can be relevant information for users and maintainers. Donaldson’s essay gives his view; it does not establish a rule binding all projects.
As an Amazon Associate I earn from qualifying purchases.
What do open-source project policies require?
There is no single requirement across the examples below. Their policies differ in scope, disclosure mechanism, and accompanying responsibilities, so check the current rules of the project receiving your contribution.
Free tools Windows power users keep installed
One-click scans. No signup required.
Linux Foundation projects
The Linux Foundation says code or content generated wholly or partly with AI can be contributed to its projects, and states that “Development and review of code generated by AI tools should be treated no differently.” Its guidance advises contributors to check that the tool’s terms do not conflict with the project’s open-source license, intellectual-property policies, or the Open Source Definition. The cited guidance does not establish a blanket AI-disclosure requirement. Read the Linux Foundation guidance.
#1 Best Overall
OpenInfra Foundation projects
OpenInfra places responsibility for a submission on the contributor and calls for review of its correctness, quality, style, security, and licensing. Its policy distinguishes Generated-By labels for generative AI contributions from Assisted-By labels for predictive AI assistance. Contributors should explain relevant context, including how much of the contribution came from the tool, and comply with any additional project-specific requirements. Read OpenInfra’s policy.
pyOpenSci peer review
pyOpenSci asks authors to disclose use of generative AI and review generated content before submission. One stated aim is to avoid making volunteer reviewers the first people to discover errors in generated material. Read the pyOpenSci policy.
How to write a useful disclosure
A useful note names the kind of work AI contributed and what human review took place. Donaldson offers this example: “Generative AI is used extensively for initial code generation and tests. All generated code is reviewed before merging.” Adapt the wording to your actual workflow; do not imply that review happened if it did not. If the destination project requires a particular label or location, use that mechanism rather than substituting a general note.
Where the project gives no required format, include enough detail for reviewers to understand the contribution. For example, distinguish generated functions or tests from predictive suggestions, and say whether the tool drafted documentation or helped with a refactor. Keep the description factual: disclosure is not a claim that the result is correct, secure, or properly licensed.
Disclosure does not replace review or responsibility
AI assistance does not transfer responsibility for submitted code to the tool or its maker. Review the result for correctness, quality, style, security, and licensing, as OpenInfra’s policy advises; pyOpenSci likewise expects authors to review generated material before peer review. A disclosure gives reviewers context, but it cannot certify the code or guarantee maintainability.
- Check the project’s current contribution and AI policies before opening a pull request or submitting a package.
- Verify that tool terms are compatible with the project’s license and intellectual-property rules.
- Understand and test the code you submit; review generated tests and documentation too.
- Use the project’s requested labels and explain the extent of assistance when asked.
What one 2026 study says—and does not say
A 2026 study in ACM Transactions on Software Engineering and Methodology reports mining 613 self-declared AI-generated code snippets and collecting 111 valid practitioner survey responses. Among those survey respondents, 76.6% said they always or sometimes self-declare AI-generated code; 23.4% said they never do. These figures describe that study’s respondents, not developers as a whole. The study reports motivations including tracking use for later review or debugging and ethical considerations; cited reasons for not declaring included substantial modification of generated code and a belief that declaration was unnecessary. Read the ACM study.
Rank #4
The study describes self-declaration practices; it does not show that undisclosed AI-written code can be reliably detected, or that disclosure itself improves trust, code quality, or maintainability. Those outcomes should not be assumed from the survey.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
