The ‘No Fakes Act’ is the proposed US legislation tackling unauthorised AI deepfakes, which was recently reintroduced for a third time in the Senate. Now it has taken a significant step towards becoming actual legislation.
The bill, which would protect “the voice and visual likenesses of individuals and creators from the proliferation of digital replicas created without their consent”, has been passed unanimously by the Senate Judiciary Committee.
This is one step further than the second version of the bill, which was reintroduced in April 2025 after the first version ran out of legislative time before the US elections in 2024.
“I hope the full Senate passes it soon so we can give Americans the security they deserve,” said senator Chris Coons, one of the four politicians behind the latest effort to get the bill passed.
As we noted in our report on its re-reintroduction recently, one change is that the bill now covers deepfakes on “streaming music platforms” – a move that has the backing of Spotify, YouTube, TikTok and all three major labels, as well as industry bodies.
Yesterday saw YouTube spark another talking point around the bill thanks to a Variety op-ed by its head of music Lyor Cohen and The Recording Academy boss Harvey Mason Jr. They welcomed the bill’s progress, but the talking point came at the end of their joint article.
“The infrastructure has to be universal. YouTube has built likeness detection technology that automatically identifies when an artist’s likeness is being used without permission,” they wrote.
“But an artist protected on one platform is still vulnerable on another. Every platform where fans can upload content needs to meet the same high bar — not just because the law might eventually demand it, but because the alternative is an ecosystem artists will eventually abandon.”
In other words, rivals (hi, TikTok, Instagram etc! But also smaller, future disrupters) should have to meet the bar set by YouTube, based on the technology that it has already built and invested heavily in.
This is an increasingly common tactic in Big Tech’s lobbying playbook. Look out for it whenever you see a big AI or social-media company welcoming or calling for specific regulations: it’s often hovering not so far below the surface.
YouTube is perfectly within its rights to try it here, but leading the charge for ‘protecting artists’ carries risks for YouTube too – inviting further questions about its approach to AI training and music uploaded to its service by independent musicians.
“Artists have to win or nothing works,” wrote Cohen and Mason in their Variety op-ed. The ‘No Fakes Act’ and YouTube’s likeness-detection tech could be wins for artists, but so would giving them the right to opt out of their work being used to train AI models.
This is a bigger theme of the big AI debates around music. Witness the way the major labels’ campaign for permission, payment and transparency from AI companies is also being applied by artist and songwriter representatives to the majors’ AI deals from their side of the equation.
If you loudly bang the drum for creators’ rights in one arena of these lobbying battles, your own arguments may come back to bite you in another area. Even so, the progress of the ‘No Fakes Act’ is good news. Let’s see if it can make the final steps now to become law.
