1,000 artists drop a silent album to fight UK's AI copyright plan

If you hit play on a newly released album and hear nothing but room tone and the faint rustle of cat paws, nothing is wrong with your speakers. It is a protest.

On Monday, a coalition of 1,000 musicians dropped a silent record titled "Is This What We Want?" The credits include heavy hitters like Kate Bush, Hans Zimmer, Annie Lennox, Damon Albarn, and Max Richter.

The project consists of 12 tracks, none of which contain music. Instead, reading the 12 song titles in order spells out a direct message: "The British government must not legalize music theft to benefit AI companies."

A record made of empty rooms and cat paws

The silence on the record is deliberate. Rather than songs, the artists submitted recordings of their empty studios and deserted performance venues, meant to symbolize what the creative world might look like if the proposed rules take effect.

Contemporary classical composer Thomas Hewitt Jones contributed ambient sound from his workspace. He noted that listeners can hear his two studio cats wandering around, which usually bother him while he composes.

The album is slated to hit streaming services on Tuesday, with proceeds and plays directed to Help Musicians, a charity organization. The protest was organized by Ed Newton-Rex, an industry insider who previously founded the AI composition startup Jukedeck, which won TechCrunch Disrupt in 2015 and was later acquired by TikTok.

Newton-Rex, who also worked at Snap and Stability AI, now runs an AI-focused nonprofit certifying companies that avoid uncredited scraping. His ongoing petition demanding fair licensing for creative works has topped 47,000 signatures, with almost 10,000 arriving in just the five weeks following the UK's latest AI strategy announcement.

The illusion of the opt-out system

The flashpoint is a planned shift in UK copyright law. To attract artificial intelligence firms to the country, officials want to allow model builders to scrape copyrighted material online for training without upfront permission or royalty payments, unless creators explicitly opt out.

In plain English, the legal default flips. Instead of tech labs asking for a license, the burden moves onto artists to chase down scrapers and tell them to back off.

Newton-Rex argues this creates an impossible bind. There is no universal registry or technical standard to opt out, nor is there any transparent way to track what datasets ingest an artist's catalog. In practice, he says, opt-out schemes are rarely adopted, effectively handing 90% to 95% of creative output straight to commercial tech companies.

Frustration has boiled over into dramatic demonstrations. Hewitt Jones recently threw a functional synthesizer into Kent harbor during an in-person protest before retrieving it broken. He said he is exploring releasing his future catalog through markets like Switzerland that provide stronger IP safeguards.

According to Newton-Rex, artists spent decades being told to publish their material online for exposure. Now, governments and tech platforms are telling them that because it was public, it is fair game. As a result, several artists have told him they are simply withholding their work from the web entirely.

Why it matters

In our view, this clash exposes the fundamental flaw of relying on an opt-out framework for AI training data.

Expecting individual musicians, writers, or illustrators to monitor thousands of scrapers is absurd. Without mandatory watermarking, auditable training logs, or standardized opt-out registries, saying artists can opt out is essentially saying they have no say at all.

For creators, this development signals that the internet's open-door era is closing. Sharing work openly online used to be the default playbook for building an audience. Now, creators have to weigh whether public distribution is worth feeding competitive models without compensation. We are already seeing the friction surface through artists withdrawing releases or seeking alternate jurisdictions.

The UK government appears eager to position itself as a friendly hub for AI labs by cutting through copyright red tape. But we think trading off an internationally renowned creative economy for marginal data concessions is a dangerous bet. If the policy alienates domestic talent and forces creators to stop sharing, tech companies will run out of fresh culture to train on anyway.