If I use copilot to generate my commit message or generate code, any evidence?

[wangsiqidahaoren]

Select Topic Area

Question

Body

I wonder if I use copilot to generate my commit message or generate code and don't declare it. Anybody will know that? Does the code generated by copilot contain a stamp?

[wangsiqidahaoren]

@mgriffin @zackfern @martinwoodward

[github-actions[bot]]

🕒 Discussion Activity Reminder 🕒

This Discussion has been labeled as dormant by an automated system for having no activity in the last 60 days. Please consider one the following actions:

1️⃣ Close as Out of Date: If the topic is no longer relevant, close the Discussion as out of date at the bottom of the page.

2️⃣ Provide More Information: Share additional details or context — or let the community know if you've found a solution on your own.

3️⃣ Mark a Reply as Answer: If your question has been answered by a reply, mark the most helpful reply as the solution.

Note: This dormant notification will only apply to Discussions with the Question label. To learn more, see our recent announcement.

Thank you for helping bring this Discussion to a resolution! 💬

[nearlythere]

I think this question is related to "watermarking" - I have also asked about this, and I am not aware of any water marking. Meaning, some kind of way to see "Copilot Used Here"--

But they do track things like "acceptance rate" - they don't show where it was accepted.

I can see WHO is using Copilot, what IDEs they use, in what organisation is the user, the numbers of lines of code shown v accepted.

So you'd think if they know the acceptance rate, they would also have the commit data. But it's not something they report on.

Check out the API documentation https://docs.github.com/en/rest/copilot/copilot-usage?apiVersion=2022-11-28

And their post about it https://resources.github.com/learn/pathways/copilot/essentials/measuring-the-impact-of-github-copilot/

[wangsiqidahaoren]

So the 'acceptance rate' is just a statistic, I can't find out whether the code snippet I specified is generated by copilot or other LLM

[nearlythere]

Yeah, I guess I don't know how to tell. I mean, the commit hash won't have anything in it, does it? Just the actual commits. So unless they are pairing the event of an accepted suggestion, with the commit hash?

I have seen research about the longevity of suggested code - meaning they can tell if suggestions end up getting CHURNED. But that was an academic paper. So, I would say this could be done? I just don't know by what means. And it's not data they share.

I would love to see water marking.