EPIC Urges FCC to Require More Disclosures Around AI Content in Political Ads—and Restrict AI Disinformation

Yesterday, EPIC filed comments to the Federal Communications Commission urging the agency to increase the granularity of information they plan on requiring broadcasters to disclose when airing political ads that use AI content.

The FCC’s proposed rule would require broadcasters to inquire into whether advertisers used AI-generated content, make general on-air announcements that an ad contains AI content, and retain records of that disclosure. However, there is a world of a difference between political ads that use AI jingles and those that use AI deepfakes of political opponents. To safeguard our elections, the public needs to know how AI is used in a political ad, not just whether AI is used.

To support the impact of the FCC’s proposed rules, EPIC further recommended requiring broadcasters to inquire into and disclose the existence of any AI watermarks within political ads. While imperfect, watermarks and other forms of AI content labels provide a way for broadcasters and the public to independently verify whether content is real or fabricated—something that will only become more important as AI-generated content becomes more convincing. Additionally, EPIC highlighted how AI disinformation functions similarly to other harmful content that the FCC has restricted in the past, such as broadcast hoaxes and news distortion. To meaningfully safeguard an informed electorate, the FCC should reconsider its decision not to restrict any uses of AI-generated content in political ads.

EPIC has filed several comments with the FCC regarding the privacy and safety of the public, and we have published two reports covering the range of harms that generative AI technologies have on our elections: Generating Harms and Generating Harms II.