AI Detection 'Journalism' Slop? The Music Industry's Biggest AI Statistic May Rest on a Black Box

One of the most widely repeated statistics in the music industry today is that nearly half of all new music uploaded to Deezer is AI-generated.

The number has appeared across industry publications, blogs, newsletters, podcasts, social media posts, conference panels, and countless discussions about the future of music. According to a recent announcement from Deezer, the company is now receiving nearly 75,000 fully AI-generated tracks per day, representing approximately 44% of all daily uploads. The figure was quickly amplified by publications including Music Business Worldwide and Billboard.

Yet remarkably few people seem interested in asking the most important question of all:

How does Deezer know?

That question may sound obvious, but it is largely absent from the coverage surrounding these claims. The discussion has instead focused almost entirely on the implications of the numbers rather than the evidence behind them.

To Billboard's credit, its reporting repeatedly describes the figures as Deezer's "claims" rather than presenting them as independently verified facts. That distinction matters. A claim is something asserted. A fact is something demonstrated. At the moment, the public evidence appears much closer to the former than the latter.

According to Deezer's own announcement, the company relies on a "patent-pending AI-music detection tool" launched in early 2025. The announcement states that the technology can detect music generated by systems such as Suno and Udio and can be adapted to identify content generated by other tools when relevant training examples are available. Deezer also states that it has filed patents related to its detection technology.

What the company does not provide is arguably more interesting than what it does.

There is no publicly available peer-reviewed paper validating the detector's accuracy. There is no published benchmark dataset. There is no publicly disclosed false positive rate. There is no publicly disclosed false negative rate. There is no independent audit demonstrating how the system performs under real-world conditions. There is no detailed technical explanation of how the detector works, what constitutes "fully AI-generated" content, or how it handles hybrid works involving both human and machine contributions.

Despite these gaps, the detector's conclusions are increasingly being treated as settled science.

That should concern anyone who values evidence-based reporting.

Imagine if a pharmaceutical company announced that 44% of patients suffered from a previously unknown condition based on an internal proprietary diagnostic tool. Journalists would immediately demand validation studies, outside reviews, methodology disclosures, and expert commentary. They would not simply repeat the figure and move on.

Yet when Deezer claims that 44% of new uploads are AI-generated, many publications appear content to cite the number and discuss its consequences without examining the underlying methodology.

The irony is difficult to ignore. Much of the music industry remains deeply skeptical of artificial intelligence. Yet when an AI-powered detection system produces dramatic numbers that support a preferred narrative, skepticism often disappears.

That is particularly strange because AI detectors themselves are not magical truth machines. They are machine-learning systems. They make probabilistic judgments based on patterns. They can make mistakes. They can generate false positives. They can generate false negatives. They can be vulnerable to changes in model architectures, new generation techniques, adversarial content, and edge cases.

The technology industry has a long history of overestimating the reliability of automated detection systems. Copyright bots have incorrectly flagged lawful content. Content ID systems have generated false claims. Spam filters have misclassified legitimate communications. Automated moderation tools have removed harmless material while allowing harmful content to remain.

Why should AI music detection be presumed immune from these challenges?

The issue becomes even more complicated when considering the modern reality of music production. Today's musicians increasingly use AI-assisted tools throughout the creative process. Producers use AI mastering systems. Engineers use AI stem separation. Songwriters experiment with AI-assisted composition tools. Vocalists rely on advanced processing technologies that blur the line between traditional production and machine assistance.

Where exactly does Deezer draw the line between human-created music and AI-generated music?

The public does not know.

The distinction matters because labels carry consequences. Deezer's own announcement states that AI-generated tracks are excluded from algorithmic recommendations and editorial playlists. The company has also stopped storing high-resolution versions of AI-detected tracks and has made its detection technology available for licensing to other organizations.

If a detection system is going to influence discoverability, monetization, visibility, and platform treatment, then its accuracy should not be treated as a secondary concern.

The industry's rush to embrace AI detection statistics has created a strange situation. Journalists routinely question government data, corporate earnings projections, polling methodologies, scientific studies, and academic research. Yet some appear willing to accept extraordinarily specific AI-detection numbers from a proprietary system without demanding equivalent levels of transparency.

Even the language surrounding these discussions deserves scrutiny. Terms such as "AI slop" have become increasingly common in music industry commentary. The phrase carries an implicit assumption that the classification is obvious and objective. But if the underlying detector remains largely opaque, then confidence in the label may exceed confidence in the evidence.

This does not mean Deezer is wrong.

The company may be directionally correct. AI-generated music may indeed be increasing rapidly. Its detector may ultimately prove highly accurate. The patents may be valid. The methodology may be sound.

But extraordinary claims require extraordinary evidence.

Right now, much of the discussion appears to rely on trust rather than verification.

The music industry frequently warns about AI-generated misinformation. It may be worth applying the same level of scrutiny to AI-generated classifications.

Before millions of songs are labeled, filtered, deprioritized, demonetized, or excluded based on AI detection systems, perhaps the industry should ask for something surprisingly simple:

The receipts.

Until then, the most widely cited AI statistic in music may tell us less about the future of music than it does about the willingness of modern journalism to repeat claims that sound authoritative without demanding the evidence necessary to prove them.