I think that it would be really impractical to try to enforce a universal ban on all AI-generated music. I think that what we should instead do is implement some toggle system so that the people who don't want to listen to music that is likely AI-generated have the option to turn off music which is probably made by AIs, rather than risking accidentally banning human producers who happen to trigger the AI detection software.

I think there's some good points that you make, but I think that you can achieve similar results not by outright banning all AI music, since it's also hard to detect with any real confidence whether the music you're listening to is AI or human-made. But you can have filtering by heuristics and use certain toggle mechanisms and labels and flag music that sounds like it might be AI generated and possibly restrict the revenue if there's a high likelihood that the music is made by an AI. But I think that an outright ban is too extreme and also not easily enforceable. Maybe the person who is putting out the music needs to go through some sort of verification system to prove that they are human. And I would be fully in support of something like that. But I think that it might be unfair to human artists whose music sounds like it could be AI generated.

Well, if you agree with me that an outright ban is not something that you can easily enforce and could potentially be unfair to human producers, then it seems that the policy proposal is not ideally formulated and that perhaps my suggestion for a toggle system would be a preferable policy proposal to the one that you're currently defending. I'm sure you could even try to have streaming services that are human exclusive, but I don't think that a government ban would be a good idea.