On July 15, Tidal stops paying royalties on fully AI-generated music, and the same move is spreading through publishing, film, science, software, and hiring at the same time. When generating anything costs nothing, the proof that a human made it becomes the product, and the only question left is who collects the fee.
⬥EPISODE NOTES⬥
Tidal is about to stop paying royalties on any track it judges to be fully machine-made. Frame that as a music story and you miss the shift underneath it. By Deezer's own detection, roughly 75,000 AI-generated tracks now arrive every day, about 44% of everything uploaded, yet that same AI music is only 1 to 3 percent of what people actually play, and around 85% of those streams are flagged as fraudulent. The flood is not an audience. It is an attack on a shared payout.
This edition follows one pattern across six industries: when the cost of generating something collapses toward zero, platforms stop paying for output and start paying for proof of human origin. Tidal cuts AI royalties. The Authors Guild sells a "Human Authored" badge for ten dollars a title. YouTube demonetizes "inauthentic" content. curl killed its bug bounty after a flood of AI slop, then reopened when the slop got good. And where no gatekeeper owns the payout, hiring, the open web, the scientific record, the flood just degrades the mechanism until no one trusts it.
In this edition of Lens Four:
🔹 Tidal's July 15 policy ends royalty attribution and direct-to-fan sales for fully AI-generated tracks, a payout decision, not a content ban.
🔹 Deezer takes in about 75,000 AI tracks a day (44% of uploads), up from roughly 10,000 a day at the start of 2025, while human uploads barely moved.
🔹 The paradox that reframes the debate: AI music is 44% of uploads but 1 to 3 percent of listening, and about 85% of those streams are fraudulent.
🔹 The first US criminal AI streaming-fraud case: Michael Smith pleaded guilty after collecting more than 8 million dollars in royalties from hundreds of thousands of AI songs and roughly 1,000 bot accounts.
🔹 curl shut down its bug bounty under a flood of AI vulnerability reports, then reopened a month later because the AI reports got good enough to read. Cutting the money did not cut the volume.
🔹 The counter-case: recruiters see about 11,000 job applications submitted to LinkedIn every minute, up 45% in a year, with no single payout to switch off.
🔹 Provenance becomes a product: Suno (2 million subscribers, about 7 million songs a day) adds identity-verified voice cloning while the Authors Guild sells human certification.
🔹 The danger tier: roughly 20% of AI-recommended software packages do not exist (slopsquatting), and close to 10% of cancer papers show paper-mill signatures.
🔹 The language turned first: Merriam-Webster made "slop" its 2025 word of the year, and YouTube quietly renamed "repetitious" content to "inauthentic."
🔹 Human filters see it clearest: DJ Sam Young asks why we need fifty versions of the same thing, and producer Gregoire Gensollen says he will remember the human moments, not the tool.
Fourth Lens: The three lenses meet at one move. Platforms re-price payouts around human origin, the market builds products that certify it, and the language teaches us to want it. That is not a defense of artists, it is a paywall around authenticity, sold as virtue, and it is arriving before audiences even asked for it. Reality has come at a premium, exactly as predicted in 2017. So the real question is not whether the real is worth more. It is this: when proof of human becomes a product, who is making the money, and who handed them the right to decide what counts as real?
▶ Read the full article and references
▶ Redefining CyberSecurity Podcast
Sean Martin, CISSP, is a cybersecurity market analyst, content strategist, and go-to-market advisor with more than 30 years of experience across engineering, product development, marketing, and media. He is co-founder of ITSPmagazine and Studio C60, host of the Redefining CyberSecurity Podcast and Music Evolves Podcast, and writes Lens Four at seanmartin.com.
Keywords: AI-generated music, Tidal, Deezer, streaming fraud, provenance, content authentication, Human Authored, Authors Guild, Suno, slopsquatting, curl bug bounty, AI slop, paper mills, AI job applications, Merriam-Webster slop, DJ Sam Young, Gregoire Gensollen, Sean Martin, Lens Four