Spotify distinguishes AI-generated artist identities: how to read its labels

Spotify is introducing labels for profiles whose public identity is a photorealistic person generated with artificial intelligence. Announced on 11 August, the measure concerns who is presented as the artist. It does not, by itself, establish how each song was made. The rollout is gradual, according to the platform’s documentation.

Two labels with different origins

Spotify’s help page distinguishes two notices. “AI Persona” means the artist has disclosed that their public identity is an AI-generated photorealistic person. “Likely AI Persona” means Spotify applied the label after reviewing the profile. The platform allows that decision to be appealed.

The distinction matters when interpreting the information: an artist’s disclosure and a platform assessment have different origins. Before sharing a conclusion, read the notice’s explanation. The category does not automatically cover every fictional character, stage name or real person who uses AI tools.

A profile does not tell the whole story of a song

AI credits answer a separate question: how the technology contributed to a recording. They can identify uses in lyrics, vocals or production. Spotify says these credits are optional, remain in beta and depend on information supplied through participating distributors.

An identity label therefore does not mean that all the music on the profile was generated with AI. Equally, the absence of credits does not prove that a song was made without these tools.

What listeners can check

Open the profile and read the explanation if a label appears. On your phone, also check the song’s available credits. The biography and official channels may add context, although a photograph or personal impression is not enough to identify an artist as AI-generated.

If these notices do not yet appear for you, remember that availability is gradual. Their absence cannot, on its own, establish who is behind an artistic identity.

Could I hear an AI-made song without knowing?

Yes. A song’s origin may not be apparent from listening, and the available information may be incomplete. That does not prove a platform deliberately recommended it because it was AI-made. It could also be in a playlist someone shared or one you selected. A virtual identity, a synthetic recording and partial AI use are different situations.

What other platforms do

Deezer says it labels tracks detected as AI-generated and excludes them from algorithmic recommendations and editorial playlists. The measure depends on what its system detects; it is not a guarantee of perfect detection.

YouTube asks music partners to declare whether content is fully, partly or not generative, and may supplement that information with other signals. This documentation describes content delivery; it does not establish an AI-free listening filter for YouTube Music.

Apple Music includes transparency data for composition, recordings and video in its provider specification. Those fields do not establish that every listener sees a notice or can block the content. An equivalent listener control was not verified in this review.

Bandcamp takes a different approach: it does not permit music or audio generated wholly or substantially with AI. This is a catalogue admission and moderation policy, rather than individual certification of every recording.

Can you choose only music by human artists?

The sources reviewed did not establish a switch that guarantees that result across these platforms. A real person may also use AI in their work: wanting human artists and wanting music without any AI involvement are different preferences.

Spotify says music from AI Persona profiles is excluded from editorial and algorithmic recommendations unless there are intentional engagement signals, such as following the artist. This policy concerns the profile’s identity; it is not a blanket block on every song made with AI.

For more control, choose albums and playlists by artists you know and check their credits. In Spotify, turn off Autoplay in playback settings to stop similar songs continuing after your selection ends. Where available, also turn off Smart Shuffle, which adds recommendations. These control selection rather than detect AI; options may vary by account and app.

What this means for the music industry

Volume and discovery are part of the challenge. Deezer reported that AI-generated deliveries exceeded half of daily uploads on peak days in June 2026. These are figures for additions to that platform’s catalogue, not half of all music listening.

As editorial analysis, a catalogue growing at this pace may increase competition for attention and make it harder for musicians to find an audience. Recommendation criteria influence both what people hear and which projects gain visibility.

Income and trust are also at stake. Deezer documents action against artificial streams and their exclusion from payments. That manipulation should be distinguished from AI creation: using the technology does not establish fraud. These figures cannot determine how much a particular artist loses either.

AI can help explore ideas and support creative processes. The communication challenge is to explain the human contribution and what a tool generated, without automatically dismissing the musician. Clear information about identity, credits and recommendation criteria helps listeners choose whom to hear and support.