Spotify set new rules on AI voice clones, spam and disclosure

What happened

On 25 September 2025, Spotify published ‘Spotify Strengthens AI Protections for Artists, Songwriters, and Producers’, announcing four policy strands: a clarified impersonation policy for unauthorised AI voice clones, a new music spam filter rolling out that autumn, support for an industry AI-disclosure standard for music credits developed through DDEX, and a recap of a beta ‘AI credits’ feature launched that April. The post states Spotify removed “over 75 million spammy tracks” in the twelve months to that date, and that total music payouts on Spotify grew from $1 billion in 2014 to $10 billion in 2024.

What the documents say

The announcement states vocal impersonation is “only allowed in music on Spotify when the impersonated artist has authorized the usage,” and that Spotify is investing in its content-mismatch review process. Spotify's living help-centre articles on AI credits and AI Persona badges, current as retrieved on 16 September 2026, add the mechanics: AI credits apply per role (lyrics, vocals, production), are submitted through a distributor and shown in Song Credits on mobile; an AI Persona badge marks an artist whose public identity, not necessarily their music, is an AI-generated photorealistic likeness, and is either self-disclosed or applied by Spotify after review. Neither document bans AI-made music, and the newsroom post states explicitly that “all music is treated equally, regardless of the tools used to make it.”

Why it matters for makers

The policy separates three distinct problems that are often collapsed into one: unauthorised likeness use (the impersonation rule), manufactured volume (the spam filter), and undisclosed method (AI credits and personas). A track can be entirely AI-generated and compliant, provided no voice is cloned without consent and no spam pattern is triggered. For a producer using a generator on a vocal, the impersonation rule is the one with teeth: it turns on whether the voice belongs to a real artist who authorised its use, not on whether AI was involved at all.

What to check before you use it

This is an editorial checklist. Before releasing a track built on a cloned or AI-modelled voice, confirm you hold the artist's authorisation in writing, since Spotify's policy makes authorisation the deciding fact, not the technology. Ask your distributor whether it supports AI-credit submission, since Spotify's own page notes only some distributors currently do. If your artist identity uses an AI-generated photorealistic image, read the Persona badge criteria before Spotify's review reaches you rather than after.

• Does the voice on your track belong to a real person, and do you hold their written authorisation to use it?
• Can your distributor submit AI credits for your release, and have you disclosed each AI-assisted role separately from human ones?
• Does your artist profile represent a real person, and if not, does it meet Spotify's definition of a photorealistic AI Persona?

Read together, the four strands describe a platform drawing lines around consent and disclosure rather than around generative technology itself, leaving the underlying question of whether a track was AI-made to voluntary, distributor-mediated credits.

Continue in Listening & research, or A DDEX-built credit standard is how Spotify's AI labels arrive.