AI isn’t creeping into music anymore. It’s already inside the room. Harvey Mason Jr., CEO of the Recording Academy and the man who runs the Grammys, now calls AI “omnipresent” in music production, according to The Verge AI’s Decoder interview. This is significant because Mason isn’t a pundit watching from the outside. He’s a legendary producer who’s worked with Janet Jackson and Beyoncé, and he says every recent session he’s been in has had AI in it.
What stands out here is the scale, and the speed it arrived at.
The number that reframes everything
How much AI music are we actually talking about? Streaming platform Deezer reports more than 50,000 AI-generated songs uploaded every single day, as detailed in The Verge AI. That’s not a niche experiment. That’s a flood.
And it’s getting harder to spot. The Verge AI notes that AI tracks are becoming tougher to identify and filter, while tools like Suno have turned into mainstream parts of how working musicians create. So the line between “made with AI” and “made by a person” isn’t blurry at the edges. It’s blurry in the middle.
The contradiction Mason has to live with
Here’s the tension. The Recording Academy’s rules say AI music isn’t eligible for the industry’s highest honors. But the CEO admits AI is now in nearly every session, including his own.
So what gets a Grammy when the tools are everywhere? That’s the question Mason has to answer, and he doesn’t pretend it’s solved. His framework, he told The Verge AI, is to gather information from trusted experts and diverse perspectives rather than rule by gut. On a topic moving this fast, that humility is probably the right call. The policy and the practice are out of sync, and closing that gap is the real work ahead.
Why this matters beyond music
There’s a line Decoder’s Nilay Patel keeps coming back to: whatever happens to the music industry happens to everything else about five years later. Music went digital first. It got disrupted by file-sharing first, by streaming first, by playlists first. Every other media business followed the same path on a delay.
If that pattern holds, the 50,000-songs-a-day problem is a preview. Publishing, film, design, and software are all heading toward the same question: when the tool is in every workflow, what counts as human work, and how do you reward it?
The business pivot underneath the AI story
There’s a second move worth flagging. The Grammys are leaving CBS after more than 50 years and moving to Disney’s ABC. Mason frames this as a content play. Disney has an appetite for documentaries, scripted projects, and music storytelling that the old setup didn’t support, and the Academy has spun up “Grammy Studios” to feed it.
Read that alongside the AI story and a strategy comes into focus. As AI commoditizes the act of making a song, the Academy is betting on the things AI can’t fake easily: live events, human stories, and a trusted brand that says “this was worth your attention.”
Practical takeaways
For practitioners and businesses watching this play out:
- Assume AI is already in your workflow. Mason’s “every session” comment is your signal. The question isn’t whether to adopt, it’s how to disclose and govern it.
- Eligibility and authenticity rules are coming. If awards, marketplaces, or platforms in your space haven’t drawn lines yet, they will. Get ahead of provenance and labeling now.
- Lean into what doesn’t scale. When output floods to 50,000 a day, scarcity moves to curation, live experience, and trust. That’s where the margin goes.
- Watch music as your five-year forecast. What the Recording Academy decides about AI eligibility is a rough draft for rules your industry will debate next.
Mason doesn’t have the final answer yet, and he’s honest about that. You can hear the full conversation at the original source.