I notice a lot of "fake” musicians shows in listing of songs and albums in recommended and similar music.
I am listening blues mostly and I receive some strange non existing singer and groups in these category like Enlly Blue, Whiskey Blues, Nina Blaze, Jerry,s Sound Room, Sienna Rose …and much more.
They are all AI music obviously.
How to fight against them?
Can Roon make something?
I’ve moved your post from Feedback > Feature Suggestions to Roon Software Discussion. I think this is an excellent topic of discussion, and could lead to a clearly articulated feature request.
I don’t want to see AI-generated content, and if the streaming services continue to add such releases to their catalogues, I’d like a way of flagging this in Roon, so it may be removed wholesale.
Whilst we can ban these artists, it would be better if this could be done another way using the Roon community (crowdsourcing.)
I think this might be an excellent initiative. I haven’t seen AI music in my Roon recommendations yet but my YouTube feed is littered with it. It would be great if Roon could develop a way to filter this stuff out or at least flag it as AI.
I had a look at one artist @Ranko_Labovic mentioned, but I’m not going to play anything since the only reason it’s there is to generate income (and take away from real artists.)
The lyrics are awful, and it is clear that a human didn’t pen them. I don’t want trash like this in Roon.
I currently use TIDAL, and it is evident that they do not have a policy for this. No idea what Qobuz does.
Qobuz are rolling out an AI detection system:
Apple have asked the labels nicely to declare AI music:
Apple’s policy seems like kind of useless given that most of this garbage is submitted via independent distributors and the people doing this are quite clearly not going to own up to doing it.
Hopefully the open source scene can develop a tool that Roon can implement to filter out this crap.
Not sure how AI-generated music can be detected. How do you screen for “poor lyrics”?
Happened upon Frost Livingston’s latest release on Qobuz, which I’d never heard of, whose sound was nothing out of the ordinary and pretty typical for the genre, imo. So, as is my usual tendency, I started googling for more information about the artist.
He/it has released five albums since September 2025 and streams on all the major services, except Bandcamp. All album artwork appears to be AI-generated. And there is no background information available, other than Frost Livingston is an independent blues-rock artist and producer who tells a real story through music, captivating listeners. These factors suggest to me that this may be AI-generated music.
The question is, how does an artist/band prove to these streaming services that the music being released is, in fact, created by humans and not AI? If truth be told, I would imagine that some of our favorite human artists are using AI now to enhance their creativity. For example, how would you feel about the Eagles, Fleetwood Mac, or Pink Floyd if you learned that one of their latest, chart-topping releases was created with the assistance of AI?
That’s a clear sign of AI-generated music, for example. I always look for signs like that first.
But without any background information, how is Roon or these streaming services to know whether Frost Livingston never had the opportunity or the intention to release his music publicly until now?
Let the community flag it in some way, so it may be hidden in Roon. Roon has previously removed poor quality material after the community flagged it, so this is doable.
If by “flag” you mean bringing it to the community’s attention via a forum post, I have no problem with it. The rest of the community can decide whether to ban it within their personal Roon app. After all, there may be some in our community who have no problem with listening to AI-generated music that appeals to their taste in music and lyrics.
But if these streaming services that claim to filter out AI content cannot develop their own systems, I think it’s a big ask of Roon to sort it all out once the music is added to TIDAL and/or Qobuz music catalog.
Wish there were a way for Roon to tag AI-generated content once we bring it to their attention, so others are made aware of what they are listening to. I do not research every new-to-me artist that Roon Radio streams my way. But it would have to be under Roon’s control to avoid inappropriate tagging by the community.
Just to demonstrate how difficult these streaming services are having detecting AI-generated music, the suspicious artist Frost Livingston made it past a highly acclaimed Deezer AI Music Detection tool per the artist’s Deezer page:
AI claims it can detect AI-generated music. So, I uploaded a track from Frost Livingston’s latest album to the free AI Song Checker V3.0, and it seems to indicate the odds of it being AI-generated aren’t 100%…
Even attorneys are having difficulty making the case…
To make things worse, AI is constantly learning from its flaws and improving on them, making it harder to detect AI-generated music in the future, in my opinion. I envision that human artists will have to get human certification to prove their music is human-made. Who said AI was taking jobs away from humans?
To Roon attention somehow, and the method would largely depend on how widespread a problem this becomes.
For instance, phony Pink Floyd albums were removed after flagging in Support > Metadata.
This is a very interesting topic. At what level would we draw the line then? 100% generated or even if just a few stems are AI?
I spent an hour exploring the AI-generated music site Suno and found quite a few of their artists’ releases on TIDAL and Qobuz. The most recently updated AI artists compiled on Zoundhub are also available on TIDAL and Qobuz.
So, I would consider it widespread and a problem to rein in if you want to completely ban AI-generated music from the streaming services Roon offers.
For now, the simplest way I can think of to handle the situation is to give a thumbs-down if Roon Radio queues up a track with AI-generated album artwork. That is usually a dead giveaway until AI figures out how to improve upon it.
My goodness - after all these years to discovers MacArthur Park was AI.
Fake music detection and labeling would be an absolutely tremendous feature. I mean this would be a truly remarkable addition - on by default!
A post was split to a new topic: AI Music Recognition Extension
How dare you! 
I love Jimmy Webb. Have you listened to Ed Sheeran’s “songs”? Now they are AI for sure. 
Whilst a couple of posts were made in jest, please do not dilute the discussion by adding further humourous content–it will be moderated.
Our preference for a particular style or genre is not the discussion point. What is, is the passing off of AI-generated content as human creativity.
If you think that’s fine, don’t complain when Marvin gets your job!
Where do we draw the line? What do we mean by AI generated? My understanding is that purely AI-generated music cannot be copyrighted, so there is a whole class of “production” or “industrial” music that requires human “editors” to modify or rearrange AI outputs to ensure the music can be legally owned and monetized. This music used to be a content production niche as background, mood for TV documentaries and YouTube creators but is now flooding streaming platforms. So here is an example of what is almost certainly a synthetic AI band, but because it also has credited human oversight, passes the Qobuz charter on fully automated uploads without human oversight. The name of the band “My Documentary” is a bit of a giveaway:
