TIDAL just drew a hard line in the sand for AI music: tracks made entirely by “text-prompted AI” will be labeled “AI,” won’t earn royalties, and can be filtered out by listeners starting July 15. The streamer also says it will remove content designed to impersonate an artist, and it will ban AI music that “exploits an individual’s or group’s music, name or likeness, deceives listeners, or diminishes the quality of our service.”

The reason this matters is simple and immediate for anyone who pays for (or depends on) music distribution: TIDAL is explicitly connecting policy to money. In its explanation, the company says it will only revoke royalties and label music that is made entirely by AI, even as it acknowledges a broader debate about the topic. Its stated priority: “ensure royalties go to original works directly produced, written, and performed by people.” That is a clear economic stance, not a vague position.

TIDAL’s move arrives as AI music use has become increasingly prominent and, frankly, harder to sort out. NME points to a study from last year showing 97 per cent of people “can’t tell the difference” between real and AI music. Around the same period, Deezer reportedly said 28 per cent of music uploaded to its platform was fully AI-generated. In parallel, there has been growing concern across the industry that people working in music could lose a quarter of their income to artificial intelligence over the next four years.

So what exactly is TIDAL doing? From July 15, it will introduce an “AI” badge on tracks found to be artificially generated. Users will also have an option to filter out all 100 per cent AI-generated content. And for impersonation, the platform promises takedowns for anything designed to impersonate an artist.

The policy text is titled “Promoting Fairness and Economic Empowerment in the Era of AI-Generated Music,” and TIDAL frames AI as “not new to music creation,” but increasingly prominent. The company draws the distinction between AI used as a tool versus AI used to fully generate a track from text prompts. It says it will revoke royalties and label only music made entirely by “text-prompted AI,” while still accepting AI-generated music generally. TIDAL justifies that acceptance with the argument that “artists should have the freedom to create with AI tools,” paired with the idea that “listeners should have the autonomy to choose the type of content they consume.”

The mid-July rollout also signals a broader platform reality: once AI tracks can be correctly identified, they no longer compete on the same terms. If users can filter out AI content, then AI-generated tracks may be pushed into a smaller lane, regardless of whether they are technically streamed. And if royalties are withheld only for 100 per cent AI-generated content, that creates a strong incentive structure. The economic consequence is not just “whether AI exists,” but who controls the boundary between tool-assisted creation and fully synthetic output.

TIDAL isn’t going solo. The source notes that Spotify previously confirmed it was removing 75 million “spammy tracks” and targeting impersonators back in September. It also references research that AI-generated songs were uploaded to dead musicians’ Spotify profiles without permission. Deezer has reportedly launched a tool to detect AI-generated music on playlists from 20 streaming platforms, which suggests the detection arms race is not limited to one company.

There is also the regulatory shadow hovering over all of this. The source points out that the government announced it would ditch “deeply damaging” plans to allow AI firms to use copyrighted works without permission. That came after Paul McCartney, Kate Bush, Dua Lipa, and Elton John led a campaign urging the government to protect artists’ work. In the same general storyline, the government had put forward controversial plans to change copyright rules and let AI firms “steal” copyrighted works without paying or seeking consent from music creators, writers, and artists.

And if you’re wondering whether this is mostly about art or mostly about risk, the last details lean toward risk management. The source says various music industry bodies worldwide penned an open letter calling for artists and songwriters to offer consent before entering any AI deals. It also mentions that famous faces have spoken out against AI music prominence, including Pope Leo XIV, SZA, Justin Hawkins, Jack Antonoff, and Billy Corgan. For platforms, the takeaway is that the conversation is not staying in the lab. It is moving into policy, enforcement, and contract economics.

For executives at streaming services, labels, and music platforms, TIDAL’s July 15 policy is a live case study in how fast the rules are changing. The “AI” badge, the ability to filter 100 per cent AI-generated content, the royalty withholding for text-prompted AI-only tracks, and the ban on impersonation are all levers that can reshape catalogs overnight. The strategic stakes are clear: whoever sets the fairest and most defensible boundary between human creation and synthetic output will shape user trust, legal exposure, and the flow of money across the entire music ecosystem.