# Plain Language Summary: DEEPFAKES Accountability Act

This bill creates new federal laws to crack down on deepfakes—fake videos, images, and audio recordings created using artificial intelligence or digital manipulation that make it look like someone said or did something they didn't actually do. The law requires anyone who creates and distributes these fake recordings to clearly label them as fake, and it creates serious consequences for people who break these rules or try to hide the fact that content is fake.

The labeling requirements depend on what type of deepfake you're dealing with. For fake videos with both audio and visuals, creators must include a spoken statement saying the content is altered, a written disclaimer at the bottom of the screen throughout the video, and a clickable link or icon indicating it's AI-generated or manipulated. For audio-only deepfakes, there must be a clear verbal statement at the beginning explaining the content is fake, with additional statements every two minutes if the audio is longer than that. For still images, a written disclaimer must appear at the bottom. The goal is to make it obvious to anyone watching or listening that they're seeing something fake.

The bill creates both criminal and civil penalties for violations. If someone knowingly fails to label a deepfake and does so with harmful intent—like creating fake sexual content to humiliate someone, making a deepfake designed to incite violence, committing fraud, or a foreign power using deepfakes to interfere in elections—they can face up to five years in prison and fines. There are also civil penalties of up to $150,000 per fake recording. Importantly, victims of harmful deepfakes can sue in federal court for damages ranging from $50,000 to $150,000 per recording, depending on the severity of harm. The bill also includes exceptions for things like parody, historical reenactments, movie productions where the original actor consented, and government use for public safety.