‘Your prompts, Spotify’s personalized picks’ - what could possibly go wrong for artists?

[Georgia Jenkins]

Home / Artists / author remuneration / Chat GPT / copyright / Generative AI / Georgia Jenkins / intermediaries / music algorithms / ‘Your prompts, Spotify’s personalized picks’ - what could possibly go wrong for artists?

‘Your prompts, Spotify’s personalized picks’ - what could possibly go wrong for artists?

 Georgia Jenkins Saturday, October 11, 2025 - Artists, author remuneration, Chat GPT, copyright, Generative AI, Georgia Jenkins, intermediaries, music algorithms

It has been a few big days for OpenAI. Grabbing headlines over last weekend as rightsholders and copyright aficionados were plunged into chaos (and, potentially gaslit?) over the release of Sora 2, ending in a pre-emptive apology video from Sam Altman, to Ninntendo, surrounded by Pokémon. However, Sora 2 is only one of a few OpenAI product announcements from their annual developer conference. Now ChatGPT has integrated a slew of third-party apps within its chatbot. From Booking.com, Expedia, Spotify, Figma, Coursera, Zillow and Canva, users can now ‘chat’ directly with apps and developers, build, without ever leaving the confines of ChatGPT.

Altman explained that:

We want ChatGPT to be a great way for people to make progress, to be more productive, more inventive, to learn faster, to do whatever they’re trying to do in their lives better. […] [Apps inside of ChatGPT] will enable a new generation of apps that are interactive, adaptive, and personalized, that you can chat with.

This GuestKat finds the integration of Spotify within ChatGPT most intriguing. Users, in some jurisdictions, can ‘request tracks, albums or playlists or ask for recommendations based on mood or genre’ within ChatGPT. Unsurprisingly, Spotify is incredibly optimistic stating that:

Spotify’s vision has always been to be everywhere you are. By bringing Spotify into ChatGPT, we’re creating a powerful new way for fans to connect with the artists and creators they love conversationally, whenever inspiration strikes.

Spotify has been quick to emphasize that the app’s integration is on a consensual, opt-in basis for users, and that no ‘music, podcasts or any other audio or visual content on [their] platform [shared] with OpenAI [will be] used for training purposes.’

However, there is room for concern.

Only a few weeks ago Spotify released AI safeguards in response to alleged artist impersonation on the platform in the face of accusations of AI generated slop and 'ghost artists'. One such ghost band, Velvet Sundown, receiving 1 million plays on Spotify and firmly placing artist exploitation within the spotlight. Their role in providing access to AI-generated content becomes even clearer if you remember their ongoing partnership with ElevenLabs that allows audiobooks recorded through AI voice software. Some have reflected, quite boldly, that: "In Spotify’s contextless vortex of passive consumption, pleasantly inoffensive AI slots right into the algorithm, infinite fodder to keep users chained to the platform.”

In light of Spotify’s integration within ChatGPT, even consensually, this context provides plenty a reason to remain cautious.

Enhanced discoverability at the continued cost of remuneration?

The headline is that Spotify’s integration into ChatGPT ‘increases’ discoverability, benefiting both artists and users. However, it is more likely that the integration of Spotify on ChaptGPT both exacerbates and perpetuates the problem of artist remuneration on music streaming platforms. ChatGPT users are ‘nudged’ towards linking their Spotify account. The result? Users potentially become more concretely tethered to algorithmic curation when requesting a playlist for Latin artists (apparently):

 

Streaming preferences have long been shaped by algorithmic recommendation systems. Algorithmic curation (e.g. personalised playlists or curated tracks for autoplay) can have a critical impact on independent artists’ livelihoods. A recent UK government study identified several issues relating to algorithmic curation (bias, opacity, and musical diversity). Perhaps most troubling being Spotify’s ‘Discovery Mode’ where rightsholders pay for promotion in exchange for a percentage of royalties. Critics reflected that it "seemed like an economic arrangement that only the most moneyed labels could afford".

Similarly, Spotify’s integration into ChatGPT has the support of Universal Music Group:

We’re supportive of these types of integrations as they offer a pathway for fans to connect with the artists they love as well as to move fluidly from discovery to enjoyment of new music—and all within a monetized ecosystem. […] While it just scratches the surface of the enormous potential of AI, it’s a positive step forward that reinforces the principle that artists must be compensated for all uses of their music on all platforms

Comment

The question that bubbles to the surface is whether this new ‘monetized ecosystem’ will reach artists. Particularly given that, in 2024, major labels reportedly made close to $15 billion in the US from streaming, and Spotify allegedly paid £7.7bn in royalties while also continuing to raise their premium subscription. Of course, the payout to artists is murkier, impacted by both the contractual relationships involving major labels, artists, and Spotify.

If anything, integrating Spotify on ChatGPT will likely entrench an increasingly inequitable system of music distribution. User prompting linked to Spotify’s algorithmic curation is a potential form of indirect monetization, propping up a system increasingly dependent on playlist exposure. One where artist inclusion, whether by a human editor or through algorithmic curation, is described as essential for continued success. And if agentic workflows are really the end goal, we might be peering at a future that has severely "diminish[ed] the intrinsic value of creative expression".

Do you want to reuse the IPKat content? Please refer to our 'Policies' section. If you have any queries or requests for permission, please get in touch with the IPKat team.