Bipartisan Senate bill takes aim at ‘deepfake’ videos

A bipartisan group of senators introduced legislation Friday to assess and cut down on the threat posed by “deepfake” videos, which are created through the use of artificial intelligence (AI) technologies to manipulate original videos.

The Deepfake Report Act is sponsored by Senate AI Caucus co-founders Rob Portman (R-Ore.) and Martin Heinrich (D-N.M.), along with Sens. Joni Ernst (R-Iowa), Brian Schatz (D-Hawaii), Cory Gardner (R-Colo.), Gary Peters (D-Mich.), and Mike Rounds (R-S.D.).

This legislation would require the Department of Homeland Security (DHS) to conduct an annual study of deepfakes and related content. It would also require DHS to assess the AI technologies used to create deepfakes and propose changes, additions to, or new regulations around these technologies.

A companion House version was also introduced Friday by Reps. Derek Kilmer (D-Wash.), Pete King (R-N.Y.), Stephanie Murphy (D-Texas), and Will Hurd (R-Texas). Hurd is a member of the House Intelligence Committee, which held a hearing earlier this month to examine the national security concerns involved with deepfakes. At the time, committee Chairman Adam Schiff(D-Calif.) described the videos as “a nightmarish scenario” to legislate.

The issue has been in the spotlight recently after a video edited to make House Speaker Nancy Pelosi (D-Calif.) appear drunk was posted online. While the video would not qualify as a deepfake, since it was slowed down to change the quality of the audio but not manipulated using AI, it showed the dangers posed to politicians’ images by manipulated videos.

The controversy was stoked by Facebook’s decision to only flag the video as fake, but not take it down. YouTube took the video down altogether.

Portman said in a statement on Friday that addressing the evolving threats posed by deepfakes will “require policymakers to grapple with important questions related to civil liberties and privacy.”