WASHINGTON – U.S. Senator Amy Klobuchar (D-MN), Chairwoman of the Senate Committee on Rules and Administration with oversight over federal elections, released the statement below following the Rules Committee passing three bipartisan bills to address the impact of artificial intelligence (AI) on our elections. The bills include the Protect Elections from Deceptive AI Act, AI Transparency in Elections Act, and Preparing Election Administrators for AI Act.
“AI can have serious consequences for our democracy and we must work with urgency to put guardrails in place. That is why we advanced three bipartisan bills out of the Rules Committee today to take on the use of AI in our elections. Whether you are a Republican or a Democrat, no one wants to see fake ads or robocalls where you cannot even tell if it’s your candidate or not, and I will continue to work across the aisle to pass these bills,” said Klobuchar.
As Chair of the Rules Committee, Senator Klobuchar has worked on a bipartisan basis to address the threat of misleading AI-generated content in our elections and safeguard our democracy.
In September 2023, Klobuchar introduced the Protect Elections from Deceptive AI Act with Senators Josh Hawley (R-MO), Ranking Member of the Judiciary Committee’s Subcommittee on Privacy, Technology and the Law; Chris Coons (D-DE), Chair of the Judiciary Committee’s Subcommittee on Intellectual Property; and Susan Collins (R-ME), Vice Chair of the Appropriations Committee and former Chair of the Committee on Homeland Security and Governmental Affairs, and joined by Senator Michael Bennet (D-CO). The bipartisan legislation would ban the use of AI to generate materially deceptive content falsely depicting federal candidates in political ads to influence federal elections.
In March 2024, Klobuchar introduced the bipartisan AI Transparency in Elections Act with Senator Lisa Murkowski (R-AK) to require disclaimers on political ads with images, audio, or video that are substantially generated by AI. The legislation requires political ads created or altered by AI to have a disclaimer, except when AI is used for only minor alterations, such as color editing, cropping, resizing, and other immaterial uses. The bill also requires the Federal Election Commission to address violations of the legislation quickly.
In March 2024, Klobuchar introduced the bipartisan Preparing Election Administrators for AI Act with Senator Collins to require the Election Assistance Commission, in consultation with the National Institute of Standards and Technology, to issue guidelines that will help election administrators address AI’s impact on election administration, cybersecurity, and election disinformation. It also requires a report on AI’s impact in the 2024 elections.
A rough transcript of Klobuchar’s full opening statement is available below. Download video here.
Senator Klobuchar: Today, we're going to consider three important bipartisan bills. This comes on a very important day, not only the day after primary day but also because this is the day that the bipartisan framework on AI was released by our colleagues, Senator Schumer and Senator Young, Senator Rounds, and Senator Heinrich. And many of us had input into the framework, practically anyone that wanted to did, and it's very, very important for the future. I like the words used to explain it – that they do it with urgency, with bipartisanship, and with humility.
It's David Brooks that once said how he had trouble writing about AI. I'll quote him exactly. He said a few months ago, “The people in AI seem to be experiencing radically different brain states all at once. I found it incredibly hard to write about AI,” he said, “because it is literally unknowable whether this technology is leading us into heaven or hell.” It can lead us into heaven in so many ways. As the Chair of this Committee and as a state which is proud to include the Mayo Clinic, I know the innovations that are possible with this. I know where we can go with this, and it's very exciting for our country, but only if we're willing to put some guardrails in place.
All of the platforms involved in this, American companies, have been very clear that we need to have some rules of the road in place. This bill that we are first considering on political deepfakes that Senator Hawley and I put together with Senator Collins and others with input from election experts. Both Democrat and Republican lawyers combed through this bill carefully, of course, because it's so important to look at it for constitutional implications and make sure it followed the Constitution, as well as for the impact that it would have in a positive way. So the fact that a number of the platforms have said yes, this is what we need to do because we can't have people tuning into TV or seeing ads or videos or get robocalls and not actually know if it's their candidate or the other. That's the first bill we're going to vote on, the Protect Elections from Deceptive AI Act. Which was introduced with Senator Coons as I mentioned, Senator Hawley is the lead Republican and Senator Collins. Senator Bennet on this Committee is also an original cosponsor.
The kinds of things we have already seen are chilling. We had, in the Republican primary, used against Donald Trump, an ad showing him hugging Dr. Fauci. It was not true. It was created by AI. There was one that got a lot of attention: robocalls made in New Hampshire that sounded exactly like Joe Biden. They were played by Senator Tillis and Senator Coons in a recent hearing in Judiciary, and you couldn't tell the difference. It was his voice telling people not to vote in the Democratic primary. That was recorded by a magician. And it just kind of goes to show you how we are going to see this resurgence of fakery and scams going on in our elections, and whether you're a Democrat or Republican, whether you're a conservative or a liberal, we cannot have our democracy undermined by ads and by videos and by robocalls when you literally don't know if it's the candidate you love or the candidate you dislike, we cannot have that happen to our democracy.
And that is why so many states have come to the rescue, but they can only do their own state ads, and their own state robocalls, and their own state videos. At least 14 states have now enacted some form of labeling so that at least the viewers of these videos know if they're real or not. If it's the real person or not. And several have looked at or adopted bans, including the state of Texas, which unanimously, in their legislature, passed a ban with the support of Governor Abbott. So to date, the states have been doing their work because they get the urgency, in the words of the authors of our bipartisan framework, they get that “they must act urgently by bipartisanship and humility.” And so, as we head into this election, I would argue the “hair on fire moment” is that we actually take this on immediately and not wait.
So the three bills are the first that I've mentioned here, which again, the states are doing different versions of this for state activity. Only on us, the Senate Rules Committee and the U.S. Senate and the U.S. House, and all these bills have or will have soon bipartisan companions. The first one I mentioned is led by Derek Kilmer, Democrat, and then Representative Gonzales of Texas. They have bipartisan counterparts in the House, and I spoke to the Speaker directly about this a few weeks ago, they are actually moving on bipartisan legislation as well. So, so far, this has not gotten into this partisan milieu, and we need to keep it there because we don't have much time left to get this done.
So we've got this Protect Elections from Deceptive AI Act, which is Senate 2770. We also have the AI Transparency in Elections Act. So I think of these as how they work together. We've got the ban bill for the most egregious example of people pretending to be the candidates. And then we have this other bill which requires labeling—disclaimers—so you actually know if it was prepared by AI. That does not cover things like changing the colors in an ad or making your hair look better, maybe hiding some of the gray. We did not include any of this in the bill, Senator Murkowski and I, because we get there's going to be practical uses of AI with cropping and the like, and we're not going to stand in the way of science. So we made this specific, and this idea actually came out of the hearing that we had with one of the Republican witnesses, that we don't want to put it on everything for every use of it, or it's going to become ridiculous to have the disclaimer.
The disclaimer will truly help in places – As Senator Collins and I were talking about this weekend – where the ban bill cannot, because of the Constitution, apply to things like satire or parody. And we've made that very clear, Senator Hawley and Senator Collins and I in our bill, but the labeling bill, the disclaimer bill is gonna really help in those kinds of incidents. So you have a parody that looks exactly like the Senator’s opponent or a Senator themselves in a political thing, at least it will say prepared using AI, in a statement on the ad. That way, at least, people won't be fooled into thinking it's the actual person. So that's the second bill.
The third bill, Senate 3897, with Senator Collins, is the Preparing Election Administrators for AI Act, which is actually, I'll get to that in a minute, but it is helping with the Election Assistance Commission and giving them some guidance on this.
So we know this is going to become one of the most significant, if not already, technological advances of our time. We know there are risks. That's why we have a major effort going on in the Senate. We want these bills or some form of them, and I'm always open to changes, as we have done with all our bills. This is the Committee that got the Electoral Count Act through the U.S. Senate, and I note, we did that and we did that on a strongly bipartisan basis. And we made changes and we got to the right place on it. I'm hopeful we can do that with these bills as well.
The bipartisan roadmap signed off by the four leaders on this includes references to all three of these bills that involve common sense rules of the road. In the last week, this Committee has heard from a bipartisan group of more than 40 national security experts and election officials calling on us to come together and advance the very bills that are before us today. That group includes former Secretary of Defense Chuck Hagel, a Republican, and Leon Panetta, a Democrat of California. Along with current and former Secretaries of State from both parties, including former Republican Secretaries Trey Grayson of Kentucky and Kim Wyman of Washington.
As I note, this is a “hair on fire” moment, and here's why AI has the potential to turbocharge the spread of disinformation and deceive voters. Like the robocalls, as I mentioned, used in New Hampshire, like the video, which was not true of President Trump hugging Dr. Fauci. Whether you are a Republican or a Democrat, no one wants to see these fake ads or robocalls. I note the states that have embarked on either ban or labeling bills include Texas, Mississippi, and Minnesota – these are ban bills – in New York, Oregon, and California. Other states that have been involved in labeling bills are Florida, Idaho, these are not exactly bright blue states, as everyone knows. Utah, Wisconsin, Michigan, New Mexico, New York, Washington. It's all over the board. It's just simply that these governors in these states said we're not going to tolerate deepfakes of our elected officials or those that are opposing them in political ads, and they were willing to stand up and say that.
Some companies are also taking action because they know that this technology has the capacity to sow chaos in elections if it is not addressed now, but we cannot rely on a patchwork of state laws and voluntary commitments. Our bill doesn't touch the state laws. Our bill is very clear that those laws will be allowed to go on. This is about federal ads. Today, we will consider the three bills. It is supported, as I noted, by experts across the spectrum, including Republican Secretary of State David Scanlan of New Hampshire, whose state, as I noted, had the deepfake robocall and that's being investigated right now of who's behind that. And he was a recent witness at a hearing. And while we must ban deep fakes, as I note, we also need disclaimers and that's the bill with Senator Murkowski.
Finally, we need to make sure those on the frontlines of our elections are prepared to respond to AI. So we're going to take up the bill with Collins to require the Election Assistance Commission to issue guidelines. That bill also includes a proposal, a very good proposal, from Senator Butler to require a report next year on AI's impact in the 2024 elections.
With more primaries this spring and summer and the general election in November, today, we have the opportunity to pass these bipartisan bills. I look forward to a productive markup. And I will now turn it over to Ranking Member Fischer. Thank you.
###
