When it comes to AI slop on the streaming platforms, we have been living in a lawless zone for a couple years — think Hamsterdam from season three of The Wire, with the drug dealers replaced by guys who do “How to Make Money with AI Music” videos. (And no Bubbles.)

So it was nice to get a glimmer of good news this week, when Apple Music introduced Transparency Tags, a new tagging policy for AI-generated songs, album art and music videos. Under the new guidelines, people submitting music to Apple can designate whether AI was used to generate a “material” portion of a track’s lyrics or music. 

The bad news is that this is only good news because the bar is so low for music streamers and AI guardrails. Transparency Tags are a weak sauce solution for three pretty obvious reasons: 

To be fair to Apple, they have said that this is only a “first step” on the way to “thoughtful policies around AI.” The natural next steps if they are serious would be to make AI tagging a requirement for all track submissions, and add teeth to the policy by banning misidentified tracks or implementing some other consequences. A crucial question: are they willing to add a layer of AI detection software, to at least spot-check what people are uploading? That would show that they aren’t trying to push this entire issue over to their customers.

None of the above is guaranteed. Tech platforms tend to resist doing moderation until they have to, and the incentives for them are strong to make as many things as “hands off the wheel” as possible. On the other hand, Apple Music likes to position themselves as the most artisanal, human-friendly music streamer, at the crossroads of technology and liberal arts. Watch this space to see if they live up to that.   

I’d love to hear your thoughts on this; email me at [email protected]