Deezer AI music detector lets users scan playlists free

Deezer launched a free AI music detector that lets users scan playlists on any major streaming service to identify likely AI-generated tracks. The web tool accepts playlist links or uploaded audio and returns a report highlighting tracks with signals consistent with synthetic music.

How the detector works

Users paste a playlist URL or upload files to the web interface, then the tool pulls audio and accompanying metadata for each track. The detector applies audio feature analysis and machine learning classifiers to look for patterns commonly associated with synthetic music production, including anomalous spectral fingerprints, repetitive microstructure, and metadata inconsistencies such as mismatched composer or publisher fields.

Output is presented as a per-track summary with a confidence score and the signals that contributed to the assessment. The tool also flags tracks where available metadata is sparse or likely erroneous, since missing or conflicting credits can indicate automated or uncredited generation. Deezer designed the interface to work with links from Spotify, Apple Music, YouTube Music and other popular services, so users can check playlists without moving files between platforms.

The company positions the detector as a consumer-facing check rather than a definitive legal determination. Confidence scores reflect the model's internal thresholds, and the interface gives users the option to inspect the specific spectral or metadata flags behind each judgment.

Limitations and context

Detection of AI-generated music remains an imperfect science. Generative audio models continue to improve and may produce outputs that increasingly resemble human recordings, which raises the risk of false negatives. Conversely, remastered or lo-fi recordings and certain production techniques can trigger false positives in automated classifiers. The tool is most reliable at surfacing suspicious items for human review rather than proving provenance beyond doubt.

Industry observers have pushed for standardized provenance and watermarking systems for synthetic audio, which would make automated detection simpler and more reliable. Without widespread watermarking, detectors must rely on heuristic and statistical signals, which vary in effectiveness across genres, recording qualities and delivery formats. The detector's performance will depend on the training data and the diversity of examples of both synthetic and human-made music the model has seen.

Privacy and data-handling are material considerations. The tool requires access to audio or playlist metadata to run analyses. Users should review the provider's stated data retention and sharing policies before scanning large or sensitive catalogs. For rights holders and curators, the detector can speed discovery of potential infringements or misattributed tracks, but it does not replace rights management, takedown procedures, or licensing audits.

Why it matters

A free, cross-platform detector lowers the barrier for listeners and curators to spot potentially synthetic tracks, increasing scrutiny of playlists and catalogs. That heightened visibility pressures platforms, rights holders and artists to adopt clearer provenance and labeling practices for generated music. For rights management and moderation teams, accessible detection tools add another layer of triage, but they do not remove the need for legal and human review.