At ARC, we have been using "AI use levels" to make it easy to disclose whether and how AI was used to create a piece of content. They were designed with mathematics in mind, but they should be useful beyond that. I am sharing them in case others find them to be a useful standard.
AI0 (no non-trivial AI use): no AI use at all, except perhaps in trivial or tangential ways, e.g. explaining background material, finding relevant literature, or rewriting prose at the level of individual phrases.
AI1 (fully human digested and composed): AI had a non-trivial role in producing mathematical ideas, but these were entirely digested by a human, who wrote everything by hand (again except for trivial writing assistance). In particular, a human actively made all decisions about how to structure the write-up, how to define terminology, how to organize proofs, and so on.
AI2 (fully human verified and edited): AI may have been heavily involved in producing both the mathematical ideas and the written content, but a human went through the content line by line, checked that it was correct and comprehensible, and made at least a minimal attempt to edit the content for clarity.
AI3 (human–AI mixture): may contain unverified AI-generated content, but a human also made non-trivial contributions.
AI4 (fully AI-generated): produced fully autonomously, except perhaps trivial edits such as re-formatting.
The distinction between AI1 and AI2 is important but somewhat subtle, so it is worth elaborating on. The key distinction is whether writing decisions were made actively or passively. A passive decision is one that was proposed by AI that a human signed off on without thinking much about alternatives. An active decision is one that either a human came up with, or an AI came up with, but a human put a reasonable amount of thought into alternatives. A piece of content is only AI1 if all writing decisions, from high-level structure to low-level notation, were made by a human actively. The simplest way to achieve this is to physically type the content without copying it from an AI-generated source.
Edited to add: suggested way to use these levels within a group of collaborators:
Require or encourage AI use level disclosure of all substantial pieces of written content.
Declare a soft norm that by default, there is no expectation for anyone to engage with any content at AI2 or above.