US senators are sounding the alarm about the threat artificial intelligence poses to elections through its ability to deceive voters. However, the prospects for a bill that could meaningfully address this problem appear uncertain.
At a hearing Wednesday, the Senate Rules Committee advanced three bills aimed at countering the threat of AI. But the only committee that could get support from Republicans on the committee would simply be to create voluntary guidelines for election officials. The bill stops short of restricting the use of AI in elections or requiring disclosure of its use, steps many states have already taken.
Minnesota Sen. Amy Klobuchar, who chairs the Rules Committee and introduced all three measures, said generative AI's ability to create deceptive images fundamentally threatens fair elections.
“We're going to see fabrication and fraud again in our elections,” Klobuchar said. “And whether you're a Democrat or a Republican, a conservative or a liberal, democracy is being undermined by ads and videos that literally don't tell you whether it's a candidate you like or a candidate you don't like. We cannot allow it to be damaged.”
Senate Majority Leader Chuck Schumer said the risks are even higher.
“If we lose all ability to distinguish between what is true and what is false, as is the case with the threat of AI, our democracy may never recover,” the New York Democrat said, calling for a hearing. said his attendance was a potential signal that he would seek to prioritize Democratic policies. law.
AI robocall, image
The dangers posed by AI were made clear in February when thousands of New Hampshire voters received robocalls from an AI-generated voice impersonating President Joe Biden urging them not to vote in the state's upcoming primary. was shown. He acknowledged that a Democratic Party operative working for his opponent had requested the call.
And last June, Florida Governor Ron DeSantis' presidential campaign used former President Donald Trump's AI to embrace Dr. Anthony Fauci, Biden's former chief medical adviser, who is highly unpopular among Republican primary voters. A video that appears to have used the generated images was released. .
Klobuchar added during the hearing that she wanted to move the issue away from a “partisan environment.” Blue, purple and red states have recently passed laws to address the threat of AI, she noted. And all three bills introduced by Klobuchar have Republican co-sponsors.
Still, there were signs that avoiding partisan politics might become impossible.
Congressional Republicans, often led by House Speaker Mike Johnson of Louisiana, have criticized previous efforts by the Biden administration to limit the spread of political disinformation by violating speech rights and giving government regulators too much power. The government has strongly opposed the proposal, arguing that it is giving too much.
Sen. Deb Fischer (R-Neb.), the ranking Republican on the committee, also spoke Wednesday about two of Klobuchar's bills: the Protecting Elections from Deceptive AI Act and the Transparency in AI in Elections Act. raised similar concerns.
The Protecting Elections from Deceptive AI Act would ban deceptive AI videos and audio about candidates for federal office. Introduced by Klobuchar in September, it has five co-sponsors, including Sen. Josh Hawley, R-Missouri.
The AI Election Transparency Act requires political ads that use AI to include a statement disclosing its use. Klobuchar introduced the bill in March and is co-sponsored by Sen. Susan Collins (R-Maine).
Fisher argued that the two bills would “increase the burden on speech” and that the definition of AI was too vague, creating uncertainty about whether speakers would be subject to penalties. . They also aim to federalize the issue, preempting state laws and usurping states' control of elections, Fisher added.
Neither bill had Republican support and passed the Rules Committee on a party-line vote on Wednesday. (Some Republicans who did not attend the hearings voted “no by proxy,” making the official vote in both cases 9-2, but this does not count as an official vote.)
The third bill, the Preparing Election Officials for AI Act, passed unanimously in committee.
This led the U.S. Election Assistance Commission to consult with the National Institute of Standards and Technology in developing voluntary guidelines for election officials on how to protect against AI threats in elections, particularly regarding the use of AI by foreign adversaries. There is a need to.
On May 13, Representatives Brian Fitzpatrick (R-Pennsylvania) and Chrissy Houlahan (D-Pennsylvania) introduced a companion bill in the House of Representatives.
Over the past year, states including Texas, Florida, New York, Michigan, Indiana, Wisconsin, Oregon, Utah, New Mexico, and Idaho have restricted the use of AI in political ads or required disclosure of AI. passed the bill.
