That's fair, allow AI slop but tag obvious AI slop as such. Hopefully they add an option to hide detected AI slop, something I wish Youtube had for instance.
I like what I see from their policy. They accept that it’s part of the industry landscape and also say it’s not monetizeable. They will likely revisit and revise their stance as things change.
I strongly agree on labeling the generated content.
> Tidal will hold AI-generated music to a higher standard of content integrity. We will not tolerate AI-generated music that exploits an individual’s or group’s music, name or likeness, deceives listeners, or diminishes the quality of our service.
I think this is a very reasonable approach, and probably also the best way to treat AI-powered copyright infringement as a whole. Just like we don't penalize artists for consuming content unless they produce actually infringing content, we should set the same focus for AI systems.
> Starting today, AI-generated music will not be monetizable. We are only in the beginning of the era of AI-generated music.
Don't really agree that this follows from the stated principle here ("... ensuring royalties go to original works produced, written and performed by people"), but will definitely help with spam etc.
Really don't mean any offence to your comment because you probably mean well but I have little tolerance for the "reasonable"/fence-sitting kind of comments on this kind of issues.
If they really cared to much about empowering people creating things for other people, like others have pointed out, they should just ban it.
Sure, in reality it's not so easy to just ban AI content because there is a spectrum of it and it's really not a clear-cut problem.
But your stance can be clear-cut, and in this messy world where there is no perfect solution one way or the other, your stance matters even more. You could either be seen as a fence sitter who allowed slop to happen, or someone who stands with human creativity battling against shitty people and their slop.
Please stop this kind of fence-sitting reasoning if you care about people.
> Tidal defines AI-generated music as music that is wholly or substantially generated by generative artificial intelligence.
I think this needs more clarity. I can think of a lot of different ways AI is used in music today as a part of the song generation process and not sure whether or not this definition would apply to it. They specifically mention developments in "text-prompted generation" but if anything that confuses the issue more, for example what about training on specific music.
This isn't a comment on how expansive or narrow the definition should be, just that they need to spell it out more to allow for consistent application (to say nothing of enforcement). If someone uses ChatGPT for lyrics, but writes the instrumentals themselves, does this policy apply? I genuinely have no idea.
They actually don't need to spell it out more. They can use discretion. This avoids hair-splitting from pedants like you because discretion is discretion and there's no argument.
I'm a Tidal subscriber, something like this is needed.
My Tidal "feed" is full of new releases that are clearly AI-generated. They use the same artist name as artists that I really like, but the music is clearly not from the artist as advertised.
I have no problem with AI-generated music, I just don't want someone trying to spoof the artists I am interested in.
I really hope someone makes a music platform in the future that is verified as human-made. Music is about connecting to human emotions, not poor facsimiles of it.
Tie it to in-person concerts and it might actually work as a business, as well as logistically – maybe the company can be a record producer in disguise and physically meet every musician they host.
As I stated elsewhere in the thread, the streaming platforms do not have the correct incentives to do this. It’s the labels that do releases that are correctly incentivized, as they need to build an audience that trusts them and enjoys the music that they release.
Independently-released music is a huge red flag. If you can’t find a single label A&R to support you, you may have to work on the quality of your output… music doesn’t exist in a vacuum. There are tens of thousands of labels across nearly every imaginable genre. Their role as gatekeeper is a valuable one.
A platform like that exists. It's called Subvert. It's a cooperative that was created as a response whenever Bandcamp has changed ownership several times. The idea of Subvert is to focus on the artists instead of the interests of the founders and investors.
> Artificial intelligence and machine learning are not new to music creation, they have just become more commonplace and advanced
In other words, there is no bright line. AI techniques have been a part of music creation from the start. What makes bad AI music hard to detect and remove is the fact that it is a much closer approximation of regular music.
The detection problem is genuinely hard. Even desktop AI agents I've been working with recently can control Spotify, fill forms, navigate apps — all indistinguishable from human interaction at the OS level. If that's hard to detect at the application layer, detecting AI-generated music at the audio layer seems like a cat and mouse game that Tidal will struggle to win without self-reporting from uploaders.
I mean you can outsource it to users, which also allows you to organically make exceptions for AI music that is actually popular.
You really just want your users, who hate AI, to see a big "REPORT AI" button they can click. The problem you are trying to solve is the perception among your users that your platform is dominated by AI slop. So at the end of the day the only thing you actually have to figure out is what your users think is AI slop, have a quick trigger on un-popular stuff, and basically never enforce on popular stuff unless there is actually some controversy.
> Starting today, AI-generated music will not be monetizable.
AKA: We will take the value, if any, AI-generated music gives, but we will not be paying royalties. This is a contradictory statement. How does the AI-generated music give value if the generated content is inherently worthless?
AI music has taken over small businesses like coffee shops and restaurants. AI music drives me nuts because, to me, it still is very much deep in an uncanny valley. That said, I can't blame the businesses because they are all (dis)incentive driven.
The music industry has stepped up its efforts globally to crack down on small businesses that play copyrighted music. They actually hire people to go into these places and spot violations.
People blame social media for the death of the monoculture but I think music rights holders have done a fair share of the damage to themselves.
I wonder if we are gonna see an emerging market where musicians are hired to provide support for AI music farms, I feel like gig musicians can easily cover/learn to play anything without much trouble
This would then become something similar to how legal tech where a license is required to practice law relies on a few lawyers sitting as a gate after the AI
I'm a musician by hobby, but used to make a living of it in my younger days. AI music has come to stay, can't do anything about it - the cat is out of the bag.
I know professional musicians that will use AI models like Suno as an aid to their tracks - mostly where they'd previously use samples or program things themselves. In these cases, where the track may be x% AI and (1-x)% Human performance, where x is very small, I think monetization or even copyright shouldn't be too difficult.
But I also know people that use tools like Suno for everything, where every single aspect of the song: Lyrics, music, production is all done by AI tools. They basically just prompt some style and vibe they want, and will upload the result. In these cases, I don't think monetization or copyright should be possible.
Then again, it is difficult to know how much AI someone used to generate their tracks, so I'm not sure how this could be enforced. I also know people that are earning very good money off their (entirely) Suno-generated tracks.
I think the question Theo asks is the most relevant: Do you think [an AI-generated song] holds as much value? And I think we've been asking this question for years now, in different contexts: recorded music (records), copyable music (dual cassette recorders, CD-burners, mp3s), sampled music (of other artists' songs), AutoTune (basically removing the human singer from the song), etc. I feel like we've been marching toward this for nearly a hundred years now, replacing the human connection to the music with something artificial.
Some of this is sensible. The copyright authority (can't recall right this moment what it is called in the US) has said only works by human beings are copyrightable. A good argument is that therefore there is no reason to pay royalties on AI generated work as it is the equivalent to public domain.
Take away the attraction to the grifters and you reduce the issue.
Of course this does not eliminate the problem of the streaming platforms tolertating AI generated work so that they do not need to pay as much out for your subscription fee.
Personally, if there were a decent Spotify alternative that had a zero tolerance to gen AI policy, I'd switch without a second thought.
Makes sense for them as a business, but still a bummer to have AI music mixed in with human music at all. To me there is literally no point to AI music. Music is communication. The artist is communicating with the listener through a pretty unique and magical asynchronous medium. AI (as we know it today) can't meet that bar and so it does not meet my definition of music.
61 comments
[ 2.6 ms ] story [ 76.7 ms ] threadI strongly agree on labeling the generated content.
Would love for YouTube to follow suit on this
> Tidal will hold AI-generated music to a higher standard of content integrity. We will not tolerate AI-generated music that exploits an individual’s or group’s music, name or likeness, deceives listeners, or diminishes the quality of our service.
I think this is a very reasonable approach, and probably also the best way to treat AI-powered copyright infringement as a whole. Just like we don't penalize artists for consuming content unless they produce actually infringing content, we should set the same focus for AI systems.
> Starting today, AI-generated music will not be monetizable. We are only in the beginning of the era of AI-generated music.
Don't really agree that this follows from the stated principle here ("... ensuring royalties go to original works produced, written and performed by people"), but will definitely help with spam etc.
If they really cared to much about empowering people creating things for other people, like others have pointed out, they should just ban it.
Sure, in reality it's not so easy to just ban AI content because there is a spectrum of it and it's really not a clear-cut problem.
But your stance can be clear-cut, and in this messy world where there is no perfect solution one way or the other, your stance matters even more. You could either be seen as a fence sitter who allowed slop to happen, or someone who stands with human creativity battling against shitty people and their slop.
Please stop this kind of fence-sitting reasoning if you care about people.
Edit: Nevermind, see riddley's comment. That's what I get for being logged in, I guess?
Tl;dr. Another one bites the dust.
I think this needs more clarity. I can think of a lot of different ways AI is used in music today as a part of the song generation process and not sure whether or not this definition would apply to it. They specifically mention developments in "text-prompted generation" but if anything that confuses the issue more, for example what about training on specific music.
This isn't a comment on how expansive or narrow the definition should be, just that they need to spell it out more to allow for consistent application (to say nothing of enforcement). If someone uses ChatGPT for lyrics, but writes the instrumentals themselves, does this policy apply? I genuinely have no idea.
My Tidal "feed" is full of new releases that are clearly AI-generated. They use the same artist name as artists that I really like, but the music is clearly not from the artist as advertised.
I have no problem with AI-generated music, I just don't want someone trying to spoof the artists I am interested in.
Tie it to in-person concerts and it might actually work as a business, as well as logistically – maybe the company can be a record producer in disguise and physically meet every musician they host.
Independently-released music is a huge red flag. If you can’t find a single label A&R to support you, you may have to work on the quality of your output… music doesn’t exist in a vacuum. There are tens of thousands of labels across nearly every imaginable genre. Their role as gatekeeper is a valuable one.
Music is about making my brain feel good. I don't care what/who makes it. If I like it, I like it.
> Artificial intelligence and machine learning are not new to music creation, they have just become more commonplace and advanced
In other words, there is no bright line. AI techniques have been a part of music creation from the start. What makes bad AI music hard to detect and remove is the fact that it is a much closer approximation of regular music.
You really just want your users, who hate AI, to see a big "REPORT AI" button they can click. The problem you are trying to solve is the perception among your users that your platform is dominated by AI slop. So at the end of the day the only thing you actually have to figure out is what your users think is AI slop, have a quick trigger on un-popular stuff, and basically never enforce on popular stuff unless there is actually some controversy.
AKA: We will take the value, if any, AI-generated music gives, but we will not be paying royalties. This is a contradictory statement. How does the AI-generated music give value if the generated content is inherently worthless?
The music industry has stepped up its efforts globally to crack down on small businesses that play copyrighted music. They actually hire people to go into these places and spot violations.
People blame social media for the death of the monoculture but I think music rights holders have done a fair share of the damage to themselves.
This would then become something similar to how legal tech where a license is required to practice law relies on a few lawyers sitting as a gate after the AI
I know professional musicians that will use AI models like Suno as an aid to their tracks - mostly where they'd previously use samples or program things themselves. In these cases, where the track may be x% AI and (1-x)% Human performance, where x is very small, I think monetization or even copyright shouldn't be too difficult.
But I also know people that use tools like Suno for everything, where every single aspect of the song: Lyrics, music, production is all done by AI tools. They basically just prompt some style and vibe they want, and will upload the result. In these cases, I don't think monetization or copyright should be possible.
Then again, it is difficult to know how much AI someone used to generate their tracks, so I'm not sure how this could be enforced. I also know people that are earning very good money off their (entirely) Suno-generated tracks.
Take away the attraction to the grifters and you reduce the issue.
Of course this does not eliminate the problem of the streaming platforms tolertating AI generated work so that they do not need to pay as much out for your subscription fee.
Personally, if there were a decent Spotify alternative that had a zero tolerance to gen AI policy, I'd switch without a second thought.