Federal lawmakers are terrified that “deepfakes” made with artificial intelligence will be weaponized to mislead people, exacerbate fraud, and create national security threats.
“Deepfakes” refer to audio, images, and video manipulated and generated by AI to appear realistic.
The House Oversight and Accountability Committee on Wednesday dug into potential problems from proliferating deepfakes, and how to stop them from upending Americans’ lives.
Rep. Nancy Mace, South Carolina Republican, said the harms from deepfakes are not hypothetical.
“Videos reportedly taken from the ground in Israel, Gaza and Ukraine have circulated rapidly around on social media only to be proven inauthentic,” Ms. Mace said at a committee hearing. “One AI-generated clip showed the president of Ukraine urging troops to put down their arms.”
Ms. Mace said she has no interest in banning all synthetic videos that others find offensive and uncomfortable, but laws cannot be enforced when fact and fiction get blurred.
Lawmakers are not the only ones with an appetite for establishing new rules for deepfakes.
President Biden signed a sweeping AI executive order last month looking to curtail potential danger from AI technology.
Mr. Biden’s team reportedly listened to former President Barack Obama in shaping the new AI order, and Mr. Obama is concerned about deepfakes.
The former president told The Verge that American society should tolerate deepfakes that lampoon politicians such as him, but new rules are needed to stop bullies from using the tech to intimidate children, such as a freshman in high school.
“We should have different rules for public figures than we do for private citizens,” Mr. Obama said in the interview published Tuesday. “We should have different rules for what is clearly political commentary and satire versus cyberbullying.”
One person’s satire is another person’s disinformation, however, and drawing clear lines will prove difficult.
The American Association of Political Consultants condemned the use of AI deepfake technology earlier this year. The association adopted a policy prohibiting the use of deepfake generative AI content as contrary to its ethical code.
Professional activists and campaign staffers may heed the industry standard, but scores of online trolls are unlikely to care, especially as the 2024 election season heats up.
Some deepfake content is already making waves on the presidential campaign trail.
Former President Donald Trump welcomed Florida Gov. Ron DeSantis to the 2024 GOP presidential race with an AI-powered video depicting the governor speaking on Elon Musk’s X with a slew of infamous characters, including Adolph Hitler and “The Devil.”
Mr. DeSantis’ team has shared a video on social media featuring images of Mr. Trump embracing Dr. Anthony Fauci that appeared manipulated, despite carrying the words “Real Life Trump.”
Congress is considering several proposals to make laws governing new AI tools. The Senate Rules Committee is zeroing in on the potential effects of AI on elections and will have a major say in any forthcoming AI legislation involving next year’s contests.
• Ryan Lovelace can be reached at rlovelace@washingtontimes.com.