Basically the title: music journalists should stop parroting Spotify’s PR claims about how they’re taking action around AI music. There are tons of examples of this, but Christianna Silva’s piece on Mashable (from one month ago) is one that caught my eye this morning. Everyone seems to have just uncritically reported on the words Spotify said, and not digging deeper at all into the actual actions Spotify has taken so far.

I realize because of Spotify’s monolithic position in the industry that even just them mouthing the words “we are doing something about AI” is somewhat mildly newsworthy, but in my experience as someone who uploaded a lot of AI music (820 songs) over the past several weeks is quite the opposite.

Silva wrote, at the end of September following Spotify’s announcement:

On Thursday, Spotify said it would start doing just that, saying in a press release that “aggressively protecting against the worst parts of Gen AI is essential to enabling its potential for artists and producers.” The platform is integrating a new spam filtering system, AI disclosures, and “improved enforcement of impersonation violations” like deepfakes.

As someone who spent years working enforcement for a platform, none of these statements give the impression of anything other than enforcing existing rules, and doing a sudden big sweep to give the public impression something is happening. It’s reputation management, imo, and little more.

Why do I say that? Because, as I said, I uploaded a huge amount of songs in a short time. In one case, I uploaded 300 AI songs in one night. There’s not even a way to label them as AI at time of upload in Distrokid, let alone surface that label in Spotify or allow users to take action on it.

All this reporting pretty much rests on taking Spotify’s word at face value, which can be problematic in journalism, as you end up whitewashing the message of others to appear more legitimate than it might otherwise seem.