That’s the difference. In practice a human has to commit fraud to do this.
But a human just using an LLM to generate code will do it accidentally. The difference is that regurgitation of training text is a documented failure mode of LLMs.
And there’s no way for the human using it to be aware it’s happening.
Are far as I know there’s one incidence of a company asserting copyright infringement against the Linux kernel, even if I’ve missed a few, it doesn’t have frequently. That will change with AI generated code, and it exposes everyone commercial entity that distributes Linux in any form to liability.
But a human just using an LLM to generate code will do it accidentally. The difference is that regurgitation of training text is a documented failure mode of LLMs.
And there’s no way for the human using it to be aware it’s happening.