The House Judiciary Subcommittee on Courts, Intellectual Property, and the Internet held a field hearing on February 2, called “Artificial Intelligence and Intellectual Property: Part II – Identify in the Age of AI.
The premise was to examine how the US Congress can support responsible innovation in applications of artificial intelligence (AI) technology and address growing concerns about the misuse of AI technology, especially with respect to the likeness, voice, and other identifying characteristics of individuals.
In that hearing, Harvey Mason Jr., CEO of the Recording Academy summarized the issue succinctly: “Recently, we saw the extreme dark side of AI fakes as some of the most famous and recognizable artists in the world have been the target of explicit, AI generated images that were circulated online.” One recent example was the circulation of pornographic deepfakes portraying singer Taylor Swift.
Testimony given by Lainey Wilson, winner of the 2024 Grammy for Best Country Album, who was excited about how artificial intelligence can be used to help people. “But I’m nervous about how it can be used to take personal rights,” she said.
She continued, “Many creators have already seen their life’s work and their own voices and images thoughtlessly ingested into AI models without their permission. … AI-generated music and video using an artists’ unique identity to perform in questionable settings or to sing lyrics they would never write or express, that doesn’t reflect who they really are, is unacceptable. It is a personal violation that threatens a person’s dignity and can put at risk everything they have worked so hard to accomplish.”
No AI Fraud, a bill before Congress
In her testimony before the Committee, Lainey Wilson spoke in support of the No AI Fraud Act, a bill pending before Congress at the time of the hearing. The text of the bill gives several recent examples of how deepfake software and artificial intelligence technologies impacted creative individuals ability to protect the use of voices and likenesses from fraudulent use; including the emulation of voices, false endorsements, non-consensual use of intimate images, and the confusion that such abuse generates.
A bill with a similar spirit has also been introduced in the US Senate, called the Disrupt Explicit Forged Images and Non-Consensual Edits Act of 2024 (aka DEFIANCE Act). One report released in 2023 found that 98% of deepfakes are pornographic. Both the No AI Fraud Act and the DEFIANCE Act are bi-partisan.
MPA weighs in
Although none of its representatives gave in-person testimony at the hearing, the Motion Picture Association submitted its own comments. Taking a position similar to those of Ms. Wilson and Mr. Mason, the MPA noted how “(a)dvancements in AI will likely lead to dramatic advancements in myriad industries from medicine to motion pictures. But with these developments also come significant concerns about potential misuses, including to create non-consensual pornographic deepfakes or spread disinformation about elections or national security matters, or otherwise deceive the public.”
In response to bills pending before Congress, the MPA cautioned that “Congress must recognize that it would be doing something that the First Amendment sharply limits: regulating speech,” saying that legislation consistent with the First Amendment is possible. “But it will take very careful drafting to accomplish the goal of addressing these harms without inadvertently chilling legitimate, constitutionally protected uses of technologies to enhance storytelling in our industry, along with other creative endeavors,” they said.
The MPA’s written testimony (linked below) went on to draft a detailed set of guidelines, meant to inform policymakers on this issue.
Statement of the Motion Picture Association, Inc. Regarding the Hearing: Artificial Intelligence and Intellectual Property: Part II – Identity in the Age of AI. Before the US House of Representatives Committee on the Judiciary Subcommittee on Courts, Intellectual Property and the Internet. Filed February 2, 2024. Motion Picture Association.
Artificial Intelligence and Intellectual Property: Part II – Identity in the Age of AI. Announcement and links to testimony. February 2, 2024. House Judiciary Subcommittee on Courts, Intellectual Property, and the Internet, US House of Representatives
No AI Fraud Act. HR 6943. A bill introduced before the US House of Representatives, 118th Congress, on January 10, 2024 by US Rep Maria Salazar (R), Florida.
The Internet is full of deepfakes, and most of them are porn. Article. by Chandra Steele. .October 18, 2023 PC Magazine
Why it matters
Quoting further from Lainey Wilson’s testimony before the Congressional committee, “It’s not just artists who need protecting – fans need it too. It’s needed for high school girls who have experienced life-altering deep fake porn using their faces; for elderly citizens convinced to hand over their life savings by a vocal clone of their grandchild in trouble. AI increasingly affects every one of us, and I’m grateful that you are considering taking action to ensure that these tools are used in responsible ways.”