Deceptive AI-generated replicas can cause real harms, such as leading the public to think an individual has endorsed a product or engaged in unsavory conduct. In many cases, however, the use of a replica—in news articles, commentary, art, and so on—is First Amendment protected speech. Existing law addresses many harms caused by deceptive images, balancing competing interests as needed. But Congress may consider targeted, proportionate fixes to fill in any gaps.

The NO FAKES Act, however, is neither targeted nor proportionate.

Congress should not reauthorize Section 702—the invasive spying program that sweeps up the emails, texts, and other online communications of millions of Americans—without significant changes to protect civil liberties.

Congress should end the many years of overreach by the U.S. Intelligence Community (IC) by requiring warrants for the collection or capture of any communication involving an American, closing warrantless “backdoor” access to data, and ensuring real transparency.