[email protected] (Richard Kenner) writes: >> > Not really. The policy could take the view that the person >> > instructing the LLM is the copyright holder because they had enough >> > human interaction with the LLM to be able to claim copyright on that basis. >> > That would provide a person to do (a) or (b). >> >> Are you sure? > > Sorry, I was trying to be careful in my wording. I meant that the policy > "could" take that view, not necessarily arguing that it *should*. [...]
Yeah, that's how I took it. >> I remember a case where, despite regular IP training, a colleague copied >> a piece of GPL code wholesale into commercial software (as part of a >> larger change). Fortunately that case was within the company's ability >> to rectify. For copyrightable changes, how could someone using an LLM >> be sure that that the LLM hadn't done the same? > > They couldn't, but I don't think that's the issue the policy is addressing. > I think it's addressing the risk that the LLM output is uncopyrightable, > not that it violates some other copyrighted file. As you pointed out, > *that* risk is not unique to LLMs. A person could also inadvertently be > infringing on a copyright or patent. Patents are an interesting case, but let's leave them aside for now :) As you say, the risk exists for code written by hand. But someone copying code from another project is at least aware that they're doing that. They might not be diligent about the licensing, or they might deliberately misrepresent the origin when submitting the code (just as someone could deliberately misrepresent their use of LLMs). But the copying itself is a deliberate, conscious action. In the situation above, with a patch co-developed by two people, the safest course is for both people to certify (a) or (b) for their portions of the work and for both people to be listed as authors. The situation I described above would then generally only happen if one person is deliberately dishonest (or at least very forgetful). In the LLM case, the situation could arise without any conscious action. There would be no deliberate dishonesty or deception. The user of the LLM simply has no idea where the code came from. So I can see the argument that the risk is different. But like you say, perhaps that wasn't what the authors of the policy were worried about. Richard
