Deezer last week launched a free online AI music detector that allows listeners on 20 of the most common streaming platforms to scan their playlists for synthetic content, marking the first time the company has made its internal detection technology directly available to the public at large. The tool, announced June 11, 2026, supports 27 languages and requires no Deezer account to use.

The release comes at a moment when the volume of AI-generated music landing on streaming services has grown at a pace the industry was not prepared to absorb. Deezer is now receiving nearly 75,000 AI-generated tracks every single day - a figure that accounts for more than 44% of the platform's total daily delivery, according to Deezer. That proportion has climbed sharply from 28% in September 2025, when the company reported detecting over 30,000 synthetic tracks daily, as documented by PPC Land at the time.

What the detector does and how it works

The tool is accessible at https://www.deezer.com/explore/ai-music-detector/ and follows a four-step process. A listener selects their streaming service, connects their account, lets Deezer scan their playlists, then views and shares the results. No subscription to Deezer is required at any step.

Underneath that simple interface sits detection technology the company has been developing since early 2025. The underlying system identifies AI-generated music by analyzing audio signals for patterns characteristic of generative models, particularly the subtle acoustic artifacts that tools like Suno and Udio leave embedded in the audio that cannot be heard by human ears but are visible in the signal data itself. Deezer applied for two patents for the technology in December 2024, focused on two distinct methods of identifying the unique signatures that distinguish synthetic content from human-recorded material.

The company has also built significant generalizability into the system. Rather than relying solely on model-specific training datasets, the detection infrastructure is designed to identify AI-generated content even without a dedicated dataset trained on a particular new model - a meaningful engineering achievement given how rapidly the generative music tooling landscape shifts.

In January 2026, Deezer licensed this same detection technology to Sacem, the French royalty collecting society, in what the company described as the first commercial deal of its kind. PPC Land reported in March 2026 on the contrast between Deezer's technical detection approach and Apple Music's self-declaration model, which Apple introduced to industry partners on March 4, 2026. Apple's system relies entirely on labels and distributors to declare AI involvement in their own submissions, with no cross-verification mechanism on Apple's side.

The data behind the announcement

The scale of the problem Deezer is attempting to address becomes apparent when looking at the platform's own data, which the company has been making public at regular intervals. In January 2025, when the detection system first came online, Deezer was seeing approximately 10,000 AI-generated tracks per day. By April 2025, that figure had already climbed to around 18% of total daily deliveries. September 2025 brought the count to 30,000 tracks daily, representing 28% of total delivery. By January 2026, the platform was receiving 60,000 AI-generated tracks per day, representing roughly 39% of total uploads. June 2026 sees the figure at 75,000 daily tracks, exceeding 44% of total daily delivery.

That trajectory is what makes the internal data point about platform switchers particularly striking. According to Deezer, 43% of people joining the platform from other streaming services already have AI music in their playlists. This figure implies that the problem is not confined to Deezer's catalog or to users actively seeking synthetic content - it is embedded in the everyday listening libraries of ordinary consumers who have no way of knowing it is there. That is precisely the gap the public-facing detector is designed to close.

Deezer commissioned a survey through Ipsos to understand consumer attitudes toward this issue. The survey, conducted in November 2025, polled 9,000 people across eight countries: the United States, Canada, Brazil, the United Kingdom, France, the Netherlands, Germany, and Japan. The results were unambiguous in their direction. 80% of respondents agreed that fully AI-generated music should be clearly labeled to listeners. 73% said they would like to know if a streaming platform is recommending 100% AI-generated music. And 52% felt that AI-generated songs should not appear in the main charts alongside human-made music.

The most arresting finding from the Ipsos survey was a blind test embedded in the methodology. When presented with two AI-generated songs and one human-made track, 97% of respondents could not correctly identify which was the real one. That single statistic frames the transparency argument more sharply than any policy document could.

"By detecting and tagging AI generated music over the past year and a half, Deezer has been at the forefront of transparency in music streaming. No other company has followed our lead yet, so we decided to make it possible for everyone to check if their playlists include synthetic music, no matter which streaming platform they use," said Alexis Lanternier, CEO, Deezer. "A vast majority of people want to know if AI music is being recommended to them and our data show that nearly half of the users joining Deezer from another platform have AI tracks in their playlists. We're expecting our AI music detector to be an eye-opening experience for listeners around the world."

AI music fraud and its royalty implications

Behind the transparency argument sits a financial one. Deezer's data shows that fully AI-generated music currently accounts for between 1% and 3% of actual streams on the platform. Yet despite being a small slice of listening activity, the primary function of these uploads appears to be fraudulent. The company found that up to 85% of the streams generated by fully AI-generated tracks in 2025 were the result of stream manipulation rather than genuine listener engagement.

When Deezer detects stream manipulation of any kind, it excludes the affected streams from royalty payments. Songs detected as AI-generated are automatically removed from algorithmic recommendations and excluded from editorial playlists, which is the company's first-line defense against synthetic content diluting the royalty pool that legitimate artists depend on.

The royalty stakes extend well beyond Deezer's subscriber base. According to a study conducted by CISAC and PMP Strategy - with participation from Deezer - nearly 25% of creators' revenues globally are at risk by 2028, an amount that could reach as much as 4 billion euros by that time. The CISAC-PMP study covered music, audiovisual, drama, literature, and visual arts, making it one of the broader assessments of AI's financial impact on the creative sector. Deezer cites this study as evidence that the challenge is systemic, not platform-specific.

The fraudulent streaming dimension of AI music has real legal precedent attached to it by 2026. A North Carolina man named Michael Smith pleaded guilty in early 2026 to federal conspiracy to commit wire fraud, having generated thousands of AI songs and streamed them billions of times through fake accounts to collect more than $8 million in royalty payments. The case, prosecuted in the Southern District of New York, illustrated exactly the mechanism Deezer's royalty exclusions are designed to block.

The competitive and industry context

Deezer describes itself as the first company in the world to detect and tag AI-generated music on its platform, and the competitive record bears that out. The company began tagging AI-generated albums with a visible label reading "content generated by AI" in June 2025, a move it described as a global first for a streaming service at the time. No comparable technical detection and labeling system had been deployed by any other major streaming platform by that date.

The contrast with Apple Music's approach is instructive. PPC Land's March 2026 coverage of Apple's AI tag announcement highlighted the fundamental architectural difference between a system that detects AI content and a system that simply invites declaration of it. Apple's technical specification describes the <ai_transparencies> container tags as optional and states that if omitted, none is assumed - meaning the absence of a tag carries no investigative consequence from Apple's side.

The broader streaming industry has moved toward AI transparency disclosures through varying mechanisms. YouTube introduced mandatory AI content disclosure requirements for creators effective May 21, 2025, as reported by PPC Land. That system places the obligation on the content creator at the point of upload, rather than on the platform's detection infrastructure. The IAB's AI Transparency and Disclosure Framework, launched in January 2026, established a risk-based disclosure model for advertising - a parallel development in a related domain where consumer trust in AI-generated material has also eroded, with Gen Z trust in AI ads falling 19 points in a two-year period.

What Deezer's public detector does that none of these initiatives do is put a diagnostic tool directly in the hands of the ordinary listener. Rather than depending on platform infrastructure, creator disclosure, or label compliance, a user can now run their own check independently of which service holds their library.

Commercial licensing and the wider ecosystem

Deezer is also making the underlying technology commercially available to other companies across the music ecosystem through a dedicated licensing channel at https://business.deezer.com/ai-detection/. The Sacem deal in January 2026 was the first disclosed commercial license. Deezer has indicated it sees the technology as an industry standard, with detection capability that covers the most prolific generative models - Suno and Udio among them - and the architectural capacity to add detection coverage for additional models as long as training data examples are available.

For marketing professionals who work in audio advertising, music licensing for branded content, or campaign measurement on streaming platforms, the scale of synthetic content in listener libraries carries direct implications. A playlist that is 43% AI-generated by track count looks different from a human-curated library in terms of listener engagement, genre distribution, and the kind of context a brand advertisement might appear within. The fact that most listeners cannot distinguish between synthetic and human-made audio - as the 97% Ipsos figure makes clear - means that campaign targeting based on playlist affinity or listening behavior could be operating on assumptions about authenticity that do not hold.

Deezer's positioning as a licensing entity for its own detection infrastructure also positions it differently from its larger rivals. Spotify, Apple Music, and YouTube all operate detection or disclosure systems built for their own platforms. Deezer is the only company to have made detection available as a standalone commercial service for third parties - while simultaneously publishing a public tool that invites users of those same rivals to audit their own libraries.

The company's approach to the 2026 World Cup context also illustrates how rapidly the synthetic content problem scales around cultural events. According to a separate Deezer announcement, 70% of unofficial 2026 World Cup anthem tracks on the platform were tagged as AI-generated as of early June 2026.

What this means for the marketing community

For media planners buying audio inventory across streaming platforms, the disclosure gap between platforms represents a measurement and brand safety consideration. An advertiser placing audio ads across 20 streaming platforms simultaneously now has access to a tool that reveals, at the playlist level, what proportion of its potential listening audience is already consuming synthetic music without knowing it. Whether that affects targeting strategy, creative briefing, or platform allocation decisions is a judgment call for individual buyers - but the data was not previously available at all.

The broader transparency question also connects to the EU AI Act's requirement that providers of generative AI systems ensure their outputs are machine-readable and detectable as artificially generated. Systems already on the market before August 2, 2026, have until December 2, 2026, to comply, according to PPC Land's coverage of the EU AI Act amendments. Deezer's approach - building detection infrastructure rather than depending on upstream declarations - positions it ahead of a compliance framework that the rest of the streaming ecosystem is still preparing to meet.

The public detector sits at the intersection of consumer rights, artist royalties, regulatory compliance, and competitive positioning. Its release today means that for the first time, any listener on any of 20 streaming platforms can run a five-minute check that the platforms themselves have not offered.

Timeline

  • January 2025 - Deezer launches its internal AI detection tool, identifying approximately 10,000 AI-generated tracks per day, representing 10% of total daily uploads
  • April 2025 - Deezer reports AI-generated tracks at 18% of total daily deliveries, or about 20,000 tracks daily; the company begins building the technical infrastructure for cross-platform compatibility
  • June 2025 - Deezer becomes the first music streaming platform to explicitly tag AI-generated music with a visible "content generated by AI" label visible to listeners on its platform
  • September 2025 - Deezer reports detecting over 30,000 AI-generated tracks daily, representing 28% of total daily delivery, as documented by PPC Land
  • November 2025 - Deezer commissions Ipsos survey across 9,000 respondents in 8 countries; over 13.4 million AI tracks detected and tagged on Deezer throughout 2025
  • December 2024 - Deezer applies for two patents for its AI detection technology, covering two distinct methods of identifying synthetic audio signatures
  • January 2026 - Deezer licenses its AI detection technology to French royalty collecting society Sacem in the first commercial deal of its kind; daily AI track delivery reaches 60,000, representing 39% of uploads
  • March 4, 2026 - Apple Music distributes a newsletter to industry partners introducing voluntary AI Transparency Tags for music content, analyzed by PPC Land
  • June 8, 2026 - Deezer reports that 70% of unofficial 2026 World Cup anthems on its platform are tagged as AI-generated
  • June 11, 2026 - Deezer launches a free public AI music detector available in 27 languages, covering 20 streaming platforms, allowing any listener to scan their playlists for synthetic content

Summary

Who: Deezer, the Paris-founded independent music streaming platform listed on Euronext Paris (ticker: DEEZR), which operates in 180 countries with a team of 550 people.

What: A free online AI music detector, available in 27 languages, that allows listeners on 20 major streaming platforms to connect their accounts and scan their playlists for AI-generated tracks. The tool is based on Deezer's proprietary detection technology, which the company has been operating internally since January 2025. Deezer is also making the underlying technology commercially available to third parties via its business licensing portal.

When: The tool was announced and launched on June 11, 2026.

Where: The tool is available globally at https://www.deezer.com/explore/ai-music-detector/ and supports 20 streaming platforms in 27 languages. The detection technology itself was developed in-house by Deezer's engineering team in Paris. Commercial licensing is available at https://business.deezer.com/ai-detection/

Why: Deezer's data shows that 43% of users arriving from other streaming platforms already carry AI-generated tracks in their playlists, most of them unaware. The platform is now receiving over 75,000 AI-generated tracks daily - more than 44% of total daily delivery - and found that up to 85% of streams generated by those tracks in 2025 were fraudulent. An Ipsos survey of 9,000 people across 8 countries found that 80% believe AI music should be labeled, 73% want to know when platforms are recommending it, and 97% cannot distinguish it from human-made music in a blind test.