DRAFT Policy
THIS IS ONLY A DRAFT - Please review this - we can edit and finalize it prior to sending it over to the OpenWrt Committer list.
These points were modified from a (ChatGPT-generated(!)) summary of the Wikipedia policy on AI-Generated content. A longer form of the discussion is at:
Wikipedia:Computer-generated content
DRAFT Policy on AI-Generated Contributions to OpenWrt
Generative AI tools are astonishingly useful for summarizing information or for creating an outline of a new document. They provide good information for the person who creates the prompt. But these LLM tools never guarantee accuracy, and their output must always be verified.
The OpenWrt Wiki and Forums strive to provide authoritative information about the project. All posts and pages must give the best information available to avoid misleading future readers. Consequently, we establish the following rules:
OpenWrt AI Content Policy:
-
Correctness: OpenWrt prohibits AI-generated content without human oversight. Contributors are responsible for ensuring that AI-generated contributions meet neutrality, verifiability, and reliable sourcing standards.
-
Disclosure Requirement: Contributors must disclose AI-generated content, particularly when substantial text is AI-assisted. For a Wiki post, the comment should include "[AI-Generated]" along with any other comments. Forum posts should state that the content was AI-generated.
-
No Copyright Violation: As with other OpenWrt content, AI-generated text must comply with OpenWrt's policies, meaning it cannot contain copyrighted material unless properly attributed.
-
Redress: As with other OpenWrt content, Admins (on the Wiki) and Moderators (on the Forum) may remove any AI-generated content that violates these rules, and warn the contributors who posted it. Contributors who repeatedly violate these rules may have their accounts removed from the Wiki/Forum.
-
Review: This policy can be reviewed, perhaps annually, to see if conditions warrant a change.