Spotify to Label AI Music, Filter Spam, and Implement Additional AI Policy Changes

Spotify has announced a series of updates to its AI policies, aiming to better label AI-generated music, reduce spam, and explicitly state that unauthorized voice cloning is not allowed on its platform.

The company plans to adopt an upcoming industry standard called DDEX to identify and tag AI music within music rights metadata. Additionally, Spotify will soon roll out a new music spam filter to catch more bad actors.

Under the DDEX system, record labels, distributors, and music partners will submit standardized AI disclosures within music rights metadata. This solution provides detailed information about AI usage—such as whether it was used for AI-generated vocals, instruments, or post-production.

"We understand that AI usage exists on a spectrum, with artists and producers incorporating AI into various parts of their creative process," said Sam Duboff, Spotify’s Head of Global Affairs, at a press briefing. "This industry standard will allow for more accurate and nuanced disclosures. It avoids forcing tracks into a false binary—where a song is either entirely AI or not AI at all," he added.

As part of the same announcement, Spotify clarified its AI personalization policy, stating that unauthorized AI voice clones, deepfakes, and any other form of sound replication or impersonation are prohibited and will be removed from the platform.

While the DDEX standard is still evolving, Spotify noted that it has received commitments from 15 record labels and distributors who plan to adopt the technology, suggesting that this could signal to others that it’s time to follow suit.

As AI tools make it easier for anyone to release music, Spotify is also introducing a new initiative to combat potential spam. This fall, the company will launch a new music spam filter aimed at detecting spammy strategies, tagging them, and then stopping the recommendation of such tracks to users.

"We know AI makes it easier for bad actors to mass-upload content, create duplicates, and manipulate search or recommendation systems using SEO tactics… we’ve been fighting these tactics for years," Duboff explained. "However, AI has accelerated these issues in more sophisticated ways, and we recognize that this requires new mitigation strategies."

Spotify plans to roll out the filters incrementally to ensure they target the correct signals, gradually expanding the system as the market evolves.

In related news, Spotify will also work with distributors to address what is known as "profile mismatches"—a scheme where someone fraudulently uploads music to another artist's profile on streaming services. The company hopes to prevent such cases before the music even reaches the platform.

Despite these changes, Spotify executives emphasized that they still support AI use in non-fraudulent contexts. "We are not here to penalize artists who use AI genuinely and responsibly. We want AI tools to help artists become more creative than ever before," said Charlie Hellman, Spotify’s Vice President and Global Head of Music. "But we are here to stop those trying to exploit the system. Only by actively preventing negative impacts can we truly benefit from the advantages AI brings," he added.

Spotify’s updates come amid the rapid growth of AI-generated music in the industry. This summer, an AI-generated band called Velvet Sundown gained popularity on the platform, prompting user complaints about the lack of transparency in labeling AI tracks. Meanwhile, streaming competitor Deezer recently reported that approximately 18%—over 20,000 tracks—of the music uploaded to its service daily is now entirely AI-generated.

Spotify hasn’t shared specific metrics on its own AI-generated content, but Duboff told reporters, "the reality is that all streaming services have nearly identical catalogs."

"People tend to upload music across all platforms," he explained, adding that uploading a track doesn’t necessarily mean someone is listening to it or that AI-generated music is profitable. "We understand that AI usage is increasingly not a binary issue, but rather a spectrum of how artists and producers are using it."