A landmark bill targeting AI deepfakes faces a US Senate Judiciary Committee vote on June 18. Five things to know about the NO FAKES Act.
In 2023, an AI-generated track called Heart on My Sleeve cloned the voices of Drake and The Weeknd, drawing hundreds of thousands of streams before it was pulled from Spotify and
YouTube. Neither artist performed on it, and cloning a voice is not clearly copyright infringement – a gap the NO FAKES Act aims to close.
The Nurture Originals, Foster Art, and Keep Entertainment Safe (NO FAKES) Act goes before the US Senate Judiciary Committee on Thursday (June 18).
It would create, for the first time in US federal This is its third attempt.
A version introduced in July 2024 ran out of time before that Congress ended, and a 2025 reintroduction stalled in the Senate Judiciary Committee as sponsors negotiated with big tech and free-speech groups warned it swept up protected speech.
A bipartisan group reintroduced the latest version on May 20.
Here are five things the music business needs to know before the vote. 1) It would create a federal right to your own voice and likeness law, an intellectual property right in a person’s voice and visual likeness, 2) Platforms could face up to $750,000 per track, 3) It builds on Tennessee’s ELVIS Act – but makes it national, 4) Its backers run from the major labels to indie artists – but not everyone is sold, 5) It lands as AI floods streaming – and detection still isn’t reliable. If the NO FAKES Act clears the committee on Thursday, it would still face the full Senate, the House, and the President‘s desk before becoming law.
Paying for Precision—The New Economics of Music Usage Data for Royalty Distribution
For decades, the music industry has accepted a compromise in how royalties from public performance are distributed. When music is played in cafes, restaurants, retail stores, gyms, and other licensed venues, businesses pay fees with the expectation that the rights creators will be compensated for those plays. But in practice, collective management organizations (CMOs) have relied on proxy data to allocate those royalties rather than concrete data. That proxy data is derived from radio airplay, surveys, and partial reporting from a limited pool of venues.This system was designed to balance cost, practicality, and precision. However, it comes with a major structural limitation—it does not measure what is actually played across the majority of physical locations. Instead, music usage and therefore royalty payments are estimated based on indirect signals. That creates a fundamental issue for musicians and rights holders who want to be paid when their music is utilized in this way. Without a benchmark tied to real-world playback in these environments, there is no way to quantify how accurate royalty distributions truly are. The industry is not working within known margins of error, but rather without a measurable understanding of how far distributions diverge from actual usage in public performance settings. This lack of precision in royalty distribution has financial consequences for everyone involved. When proxy-based systems are used, value is redistributed along blurred lines. Some rights holders benefit disproportionately because their music is over-represented in proxy datasets. Others are underpaid because their music, while played in venues, is not captured. This is particularly relevant for independent artists and niche genres that may perform strongly in public spaces, but lacks broader airplay data. Other sectors outside the realm of music routinely invest in better data to improve outcomes. Music royalty distribution for public performance is now approaching the same inflection point. The central issue is no longer technological feasibility, but whether the industry is willing to move beyond systems that are not capable of measuring their own accuracy. As expectations around fairness and transparency increase, reliance on proxy data becomes harder to defend.
Is Anyone’s Music Safe? Newly Identified ‘Giant Datasets’ Containing Millions of Songs Raise Fresh Questions About Music AI Training Processes
Are gen AI companies actively developing their music models with the same collections of copyrighted tracks? And despite ongoing discussions about free-for-all training, is this process far more systematic than we’ve been led to believe?
These and other pressing questions are taking center stage following an investigative report from The Atlantic’s Alex Reisner, who identified “four giant datasets of songs that are being shared within the AI-development community.” Unsurprisingly, even in light of the noted report, we don’t have a concrete answer. Said report pinpointed four training datasets consisting of north of 22 million recordings between them – including two collections clocking in at closer to 100,000 recordings apiece, one containing 9.7 million songs, and the last with roughly 12.3 million tracks. Google and Stability AI have reportedly utilized tracks from one of the 100,000-song datasets, the Free Music Archive. Owing to “the industry’s secrecy around training data, we don’t currently know who has used the others” – though all four are said to have been “downloaded thousands of times” in total, per the report.
After Spotify Eyes Concert Streaming, YouTube Launches ‘Music Nights’ Exclusive Live Concert Series
Just a week after Spotify reportedly began shopping around to secure licensing to stream live events, YouTube claps back with “Music Nights.” The new series of exclusive concerts designed “for dedicated fans” will include release parties, intimate shows, and special tour stops, with the first three to feature Isaiah Rashad, Kacey Musgraves, and Bleachers.
“This year, we’re hosting Music Nights in music hubs across the globe, from Los Angeles, New York, Paris, London, and Tokyo to unique destinations with a special meaning to artists, like New Braunfels, Texas, and Asbury Park, New Jersey,” reads YouTube’s blog post. “You can dive into the full performances, relive standout tracks on repeat, and explore exclusive behind-the-scenes moments on Shorts, directly on each artist’s Official Artist Channel.” YouTube’s Music Nights is just the latest move in the ongoing rivalry between YouTube Music and Spotify. The two companies have become increasingly competitive in the podcast arena in recent months. But the live music scene is already fiercely competitive, and YouTube has a well-established foot in that door—which might explain the timing of the Music Nights announcement.
Data Drain: The Land and Water Impacts of the AI Boom
A low hum emerges from within a vast, dimly lit tomb, whose occupant devours energy and water with a voracious, inhuman appetite. The beige, boxy data center is a vampire of sorts—pallid, immortal, thirsty. Sheltered from sunlight, active all night. And much like a vampire, at least according to folkloric tradition, it can only enter a place if it’s been invited inside.
In states and counties across the US, lawmakers aren’t just opening the door for these metaphorical, mechanical monsters. They’re actively luring them in, with tax breaks and other incentives, eager to lay claim to new municipal revenues and a piece of the explosive growth surrounding artificial intelligence.
That may sound hyperbolic, but data centers truly are resource-ravenous. Even a mid-sized data center consumes as much water as a small town, while larger ones require up to 5 million gallons of water every day—as much as a city of 50,000 people. Powering and cooling their rows of server stacks also takes an astonishing amount of electricity. A conventional data center—think cloud storage for your work documents or streaming videos—draws as much electricity as 10,000 to 25,000 households, according to the International Energy Agency. But a newer, AI-focused “hyperscale” data center can use as much power as 100,000 homes or more. Early in the AI boom, in 2023, US data centers consumed 176 terawatt-hours of electricity, roughly as much as the entire nation of Ireland (whose electric grid is itself nearly maxed out, prompting data centers there to use polluting off-grid generators), and that’s expected to double or even triple as soon as 2028.
(ie: For every question asked CHAT GPT, 1/2 liter of water
is consumed at 2.5 billion requests per day)
Suno’s Legal Battle Against Sony Music and UMG Just Got 100 Times More Serious—Literally
Universal Music and Sony Music are dramatically expanding their litigation against AI music giant Suno, claiming over 61,000 copyright infringements.
Just moments after Sony Music Entertainment expanded its lawsuit against AI music company Udio, Sony and Universal Music Group dramatically expanded their litigation against Suno, the biggest AI music platform in the game. Instead of just 560 works, the music label giants are claiming infringement of over 61,000 works—at least, if a judge approves their latest amended complaint.
In both of these expanded cases, the labels used Audible Magic, an industry-standard audio fingerprinting technology, to scan Suno’s training data, confirming that the platform used “millions” of their copyrighted tracks to train its AI models. Now, that data source is being submitted into the court record to await approval. Naturally, Suno strongly opposes this move, arguing it would effectively reset the case and delay their ability to pursue their “fair use” defense in a timely manner. However, the labels state that they could settle that matter via summary judgment separately before completing the discovery required for the newly submitted 61,000 tracks.
Musicians’ Union Sues Major Labels for Artists’ Share of AI Song Generator Settlement Money
The The American Federation of Musicians is suing major record companies Universal Music Group and Warner Music Group over the labels’ recent moves to settle their lawsuits with AI music generators Suno and Udio, arguing that the settlements’ benefits aren’t reaching the musicians themselves.
“While the Defendants protected their own interests and created a significant source of new revenue with the retrospective settlements and prospective licenses, they have refused to compensate the musicians whose work – created with their own instruments and through their talent, creativity, and hard work – is fed into AI machines for profit,” the AFM said in the complaint filed in federal court. The AFM’s lawsuit comes months after UMG and WMG reached settlements with Udio and Suno last fall. UMG, the world’s largest music company, struck the first deal, announcing a settlement and partnership with Udio in late October of 2025. WMG came after, announcing a partnership of its own with Udio in mid-November. Weeks after that, WMG became the first (and so far the only) major label to settle with
Suno. The “big three” record companies, which includes Sony Music Group alongside UMG and WMG, first sued Suno and Udio in 2024, accusing that the AI music generators of massive copyright infringement by training their models on thousands of iconic songs without permission. Sony is the lone major music company that hasn’t settled with either AI company.
Spotify's AI bet: more of everything, less of what you want
Spotify was a music app at one time. Then it added podcasts. Then audiobooks. Now the company is piling AI features into its app at a pace that can feel overwhelming. The latest wave, announced at its investor day, skews heavily toward using AI to generate content rather than using AI to help users find content they actually want.
Until now, Spotify has been largely a platform for human-created content — music, podcasts, and audiobooks. As it adds AI-powered tools to generate all of those formats, the app is poised to look very different. That shift is also creating friction — AI can now produce music faster than Spotify can manage it. The company is no longer focused solely on consumption — it’s actively nudging users to create content, too, even if it’s just for themselves. The risk is that this trades depth for breadth: The more time users spend making sense of a cluttered app, the less time they spend discovering and listening to content by other creators. This raises the question: Is Spotify deepening its competitive moat or diluting what made it essential?
