As a music and intellectual property attorney, I do not view AI-generated deepfakes as a future problem. I have already seen AI-generated websites using a well-known author’s name, image, and recreated voice to generate content that appears authentic. It wasn’t.
What once required a production studio can now be done by virtually anyone with access to AI tools.
The harm is no longer hypothetical. On June 22, the news wires announced a deal between Getty Images and OpenAI to use Getty Images’ library of millions of photos on OpenAI’s platform. Getty’s stock popped on the news.
What rules, if any, will govern use of the millions of celebrity images in Getty?
Last week, the Senate Judiciary Committee unanimously advanced the bipartisan NO FAKES Act, legislation that would create federal protections against unauthorized AI-generated replicas of a person’s voice and likeness.
For decades, artists, songwriters, authors, performers, publishers, and content creators have relied on copyright, trademark, and right-of-publicity laws to protect the value of their creative work and identities.
Today, artificial intelligence presents a new challenge unlike anything we have seen before.
WHAT IS THE NO FAKES ACT?
The Nurture Originals, Foster Art, and Keep Entertainment Safe (NO FAKES) Act would create a federal property right protecting every American’s voice and visual likeness from unauthorized AI-generated replicas.
The bill would allow individuals to authorize—or prohibit—the creation and distribution of AI-generated replicas of their voice or likeness.
Importantly, the legislation includes protections for legitimate First Amendment activities, including news reporting, commentary, criticism, parody, satire, and other protected forms of expression. (And if it passes, we’ll spend decades litigating what’s parody or fair use)
WHY THE CREATIVE INDUSTRIES SUPPORT THE BILL
The legislation has attracted support from an unusually broad coalition including Universal Music Group, Sony Music Entertainment, Warner Music Group, Spotify, Google, OpenAI, IBM, and YouTube.
Historically, technology companies and content owners have often found themselves on opposite sides of intellectual property debates. Here, however, many appear to agree that unauthorized AI replicas pose a serious threat to creators and consumers alike.
For musicians, authors, publishers, performers, and content creators, unauthorized AI replicas threaten reputations, goodwill, and the economic value of creative works.
WHY THIS MATTERS?
As both a trademark attorney and a music attorney who has spent more than thirty years fighting for the rights of artists, composers, authors, publishers, entertainment companies, and brand owners, I have watched technology repeatedly outpace the law.
The fundamental issue raised by AI deepfakes is ownership of personal attributes and creative intelligence when used for profit.
And there’s a bigger issue, one troubling almost everyone.
Who and what can we trust? What’s real and how do we know?
Who owns your voice? You do, but who makes $ from it and who controls the right to say yes or no?
Who owns your likeness? You do, but who is using it to market without permission?
Who controls the commercial use of your identity? We’ll soon find out.
For most creators, the answer should be obvious.
WHY I CARE ABOUT THIS ISSUE
Throughout my career, I have represented artists, composers, authors, publishers, entertainment companies, and brand owners.
Today, I am seeing something entirely new: AI systems capable of replicating a person’s voice, likeness, and identity with startling realism.
They aren’t.
And consent may not be required. Ultimately, artists and performers may be required to forgo individual consent and settle for a broader licensing scheme that allows use, with the right to participate in a royalty pool. Not ideal, but that precedent was set in the years of fighting over streaming rights and royalties. It’s the easy way out.
These issues are no longer theoretical. They are already affecting creators, publishers, and estates.
The debate is always the same. Technology promises innovation while creators fight to preserve the value of their work, identities, and livelihoods. It happened with YouTube and pirated uploads to streaming services. It took years for Google to finally start paying.
Innovation is important and drives our economy and society forward.
But so are rights in creative works.
Performers should not wake up to find AI-generated versions of themselves circulating online without permission.
That is why the NO FAKES Act matters.
THE OPPOSITION WAS PREDICTABLE
Not everyone supports the NO FAKES Act. The Electronic Frontier Foundation has urged Congress to reject the bill, arguing that it could chill free speech and sweep too broadly. Frankly, that opposition comes as no surprise.
That’s the same position they’ve taken for a couple of decades now.
The NO FAKES Act is not about suppressing legitimate speech, criticism, parody, or news reporting. The bill specifically protects those activities.
It is about preventing the unauthorized commercial exploitation of a person’s identity through technology that did not exist when most right-of-publicity laws were written.
There is a fundamental difference between free expression and appropriating someone else’s voice, likeness, reputation, or goodwill for your own benefit.
Creators deserve to control their own identities.
That principle should not be controversial.
WILL CONGRESS ACTUALLY ACT?
For once, I believe the answer may be yes.
The unanimous vote in the Senate Judiciary Committee is significant. The broad support from both technology companies and the creative industries is significant.
But after decades of watching artists, authors, composers, performers, and publishers fight to protect their work against new forms of technological exploitation, I am more optimistic than I have been in years.
Congress finally appears to understand that a person’s voice and likeness are valuable property rights deserving meaningful protection.
The technology has arrived.
The abuse is already occurring.
Now Congress must decide whether the law will finally catch up.
Do you believe AI companies should be required to obtain permission before replicating someone’s voice or likeness?