Deezer launches free tool to detect AI-generated music on Spotify and Apple Music

What happened?

Deezer, the French streaming platform, has launched a free tool that allows users to scan playlists from rival services like Spotify and Apple Music to detect songs generated by artificial intelligence. According to Gizmodo, the tool analyzes metadata and sound patterns to identify tracks likely created by generative AI. This move comes amid a context where AI-generated music has saturated platforms: it is estimated that Spotify removed more than 7% of songs uploaded in 2023 due to suspected synthetic content, and Universal Music Group has asked streaming services to block scraping of their catalogs to train AI models. Deezer, which has 16 million subscribers compared to Spotify's 220 million, seeks to differentiate itself through transparency, a strategy it already used in 2022 by introducing active listening-based payments to combat stream fraud.

Why is it important?

The proliferation of AI-generated music has been a hot topic in the industry. In April 2023, a fake song attributed to Drake and The Weeknd, 'Heart on My Sleeve', accumulated millions of streams before being removed. This incident highlighted platforms' inability to proactively detect synthetic content. Deezer, which competes directly with these giants, bets on transparency as a differentiator. The tool not only benefits listeners who want to avoid synthetic content but also pressures Spotify and Apple Music to adopt similar measures. Additionally, it addresses an economic problem: according to a report by the British Phonographic Industry (BPI), AI-generated music could cost human artists up to $1.2 billion in royalties by 2028 if left unregulated. Deezer's approach also contrasts with other platforms: YouTube has required AI labeling since November 2023, but only for realistic content; TikTok has done so since March 2024. Deezer, however, goes a step further by allowing scanning of competitors' playlists.

Consequences for the industry and users

This measure could accelerate the adoption of AI labeling standards across the industry. If Deezer succeeds in making users trust its platform more by guaranteeing authenticity, other services may be forced to follow suit. For artists, it represents a tool to protect their work from unauthorized imitations. However, it also opens debates about the definition of 'AI-generated music'. For example, should songs that use AI for mastering or composing parts be labeled? Deezer's tool is based on a model trained with data from its own detection system, which has already identified over 10,000 suspicious tracks in its catalog. Nevertheless, the technology has limitations: it can produce false positives (human songs flagged as AI) and false negatives, especially with advanced AI that mimics human styles. Market reaction has been mixed. Spotify has not officially commented, but internal sources cited by Music Business Worldwide indicate they are developing their own detection tool. Apple Music, for its part, has updated its terms to explicitly prohibit unlabeled AI-generated content.

What should readers know?

• The tool is free and does not require a Deezer account to scan public playlists from Spotify and Apple Music, although logging into Deezer is required to use it.
• It is not perfect: Deezer admits there may be false positives (human songs flagged as AI) and false negatives, especially with advanced AI that mimics human styles. Reported accuracy is 85% in internal tests.
• Deezer plans to integrate this technology into its own platform to automatically label suspicious content, following the model of YouTube and TikTok.
• It is part of a larger trend: platforms like YouTube and TikTok already implement AI labels, and the European Union is considering including AI-generated music in its Digital Services Act.

“Transparency is key to maintaining trust in digital music,” said a Deezer spokesperson. “We want listeners to know exactly what they are listening to.”

In summary, Deezer's tool is a step toward a more honest music ecosystem, but its long-term effectiveness will depend on collaboration across the industry. If other platforms adopt similar standards, we could see universal AI labeling in streaming. However, the technical and legal challenge is enormous: defining what 'AI' means in music, avoiding censorship of legitimate uses, and ensuring detection is not discriminatory. For now, Deezer has struck the first blow in a war that is just beginning.