What should we do with CLs generated by AI?

[Ian Lance Taylor]

I happened to notice that https://go.dev/cl/741504 has a line
"Co-Authored-By: Claude Opus 4.5 <nor...@anthropic.com>". (I may have
missed this in other CLs.)

Do we have a policy about accepting CLs that were co-written by AI? In
particular, are there any copyright concerns? Has Anthropic signed the
copyright disclaimer? Is that even meaningful given that Claude Opus
has undoubtedly been trained on software distributed under licenses
that are not compatible with Go's license?

These are questions across the broader free software ecosystem, of
course. If anybody is aware of any official stance taken by other free
software projects, that would be helpful.

Thanks.

Ian

[Robert Griesemer]

Good questions, indeed.

With respect to this specific CL, it seems to me that this line in question (in the CL description)

could just be removed, couldn't it? Or does Claude use require that its artifacts be attributed?

If the latter, Claude use in general would have a much bigger problem.

For comparison, I use a spell checker all the time, and now Google also provides me with better phrasing/prose,

yet I am not aware that people customarily mention that their texts were proofread by an AI, or that one is required

to do so. (I can see that it may be the "correct" thing to do when writing a longer article that was significantly

shaped in content by an AI.)

- gri

--
You received this message because you are subscribed to the Google Groups "golang-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-dev+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/golang-dev/CAOyqgcVYa-AfmgF%3DmgOshhmKNnjxHdWzMnvZ1rYEUqo%2BPR%3D6fg%40mail.gmail.com.

[Keith Randall]

The lawyers at Google provided this helpful (?) text:

All submissions to Google Open Source projects need to follow Google’s Contributor License Agreement (CLA), which covers any original work of authorship included in the submission. This doesn’t prohibit the use of coding assistance tools, including tool-, AI-, or machine-generated code, as long as these submissions abide by the CLA's requirements.”

Leaving it up to "abiding by the CLA's requirements" as to whether it is ok.

To view this discussion visit https://groups.google.com/d/msgid/golang-dev/CAKy0tf5DsWPNhjsf9_HEgxXt4DTkGoLbetCuVgDftnYjcM3ksA%40mail.gmail.com.

[Keith Randall]

I suspect "abiding by the CLA's requirements" may just mean "I own the copyright of what I am submitting", or at least "no one else but me owns the copyright of what I am submitting". Which may be true for AI-generated code?

[John S]

The truth is, you don't know how or on what it has been trained, so you can't assume anything in that aspect.

I think that ultimately, it is the comitter responsibility to make sure that the code provided abides to the licensing scheme of the project.

[Rob Pike]

Just exactly _what_ was written by the AI? That matters. The code? The CL description? And who has responsibility for maintenance.

This is a very slippery slope. Be careful on your first step. I recommend simply saying, no.

-rob

--

You received this message because you are subscribed to the Google Groups "golang-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-dev+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/golang-dev/CAAZVx9-N06zXkQUZioAQgGBa0fsANXM0Y%3DqwxNH9iYiBGBGQYg%40mail.gmail.com.

[robert engels]

I agree. I think there are going to be massive lawsuits by open source developers and orgs. That will cripple AI - or they will pay. This isn’t fair use - it’s stealing. The purpose of fair use was to encourage building upon - free - which isn’t what AI agents do. 

On Feb 3, 2026, at 5:17 PM, Rob Pike <r...@golang.org> wrote:



To view this discussion visit https://groups.google.com/d/msgid/golang-dev/CAOXNBZQtbwBPXza3Dk32xKHQ39a04fijks86fKmcdEcZ3YERGA%40mail.gmail.com.

[Alan Donovan]

r wrote:

> This is a very slippery slope. Be careful on your first step. I recommend simply saying, no.

I suspect a significant fraction of CLs we receive today already include fragments of LLM-generated code, whether the authors admit it or not. If they have thoroughly analyzed and reworked it into the context of the CL, as they might in times past have incorporated a raw snippet of code from Stack Overflow before smoothing it out, then the result is currently protected, at least according to the guidance in Section F of https://copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf. But I suspect there is more legal excitement ahead. Slippery indeed.

On Tue, 3 Feb 2026 at 21:53, 'robert engels' via golang-dev <golan...@googlegroups.com> wrote:

I agree. I think there are going to be massive lawsuits by open source developers and orgs. That will cripple AI - or they will pay. This isn’t fair use - it’s stealing. The purpose of fair use was to encourage building upon - free - which isn’t what AI agents do. 

I agree, but the outcome of Bartz v. Anthropic was that they were expected to pay for one, not zero, copies of each book they ingested yet were otherwise free to keep doing what they do. I'm curious what will happen in NYTimes v. MS & OpenAI.

[robert engels]

I agree that the Bartz case is problematic. But I most open source project licenses are more stringent - especially in terms of attribution or derivative works. It’s going to be interesting indeed. 

On Feb 3, 2026, at 10:23 PM, Alan Donovan <al...@alandonovan.net> wrote: