The generative AI-assisted fake president was convincing enough to alarm members of Congress who’ve been asleep at the digital wheel, but, now that we’re in the throes of the 2024 election, passing any election related measures has only gotten more difficult, which means this year’s real election may prove to be the fakest contest in history.
Realistically faking the president’s voice was surely a wake-up call in the Senate.
“I think all of us paid a lot of attention to that and we have members who've been warning about it for a while,” Sen. Tom Kaine (D-VA) told Raw Story at the Capitol this week.
Senators took notice, but that doesn’t mean senators are acting.
“I haven't heard discussion of Senate action about it,” Kaine said.
New Hampshire officials are investigating who made the Biden deepfake and whether it was an illegal attempt to depress voter turnout using artificial intelligence. By definition, a deepfake is an “image or recording that has been convincingly altered and manipulated to misrepresent someone as doing or saying something that was not actually done or said,” according to Merriam-Webster.
While the Biden deepfake could be the handiwork of a foreign actor — think Russia or even North Korea — generative AI platforms such as Chat GTP are now accessible to all people. Both major political parties would be foolish not to be experimenting with how the new tech, which is revolutionizing everything from wars to the workplace, can give them an electoral assist.
But with that opportunity comes great risk. Republican presidential frontrunner Donald Trump learned that in December, when a shady, scammy business entity used an almost pitch-perfect replica of the former president’s voice to hawk Trump-themed — and unauthorized — “gold bars,” Raw Story revealed.
And that is what’s complicating congressional action.
ALSO READ: Racism, fascism and cruelty: Donald Trump’s New Hampshire performance in nine quotes
“There's plenty of impetus to act, the problem is that people will act according to what's in their interests, you know, at the moment. So is there a broader impetus to act? I don’t know,” Sen. Kevin Cramer (R-ND) told Raw Story. “I don't think anybody likes it.”
Last year, Senate Majority Leader Chuck Schumer (D-NY) tried to compel senators to focus on generative AI. Schumer and a bipartisan group of senators even hosted rare all-Senate briefings that tackled AI issues. (The briefings were closed-press — and thus, closed to the public.
After those briefings — which included one classified briefing — Schumer and Co. hosted a series of AI forums, which were also closed to the press corps and public, even as the likes of Elon Musk, Mark Zuckerberg, Bill Gates and other titans of tech were, quite literally, seated at the head table — or dais, such as it was. Senators, meanwhile, sat in the audience looking more like middle schoolers than policy makers (at least for the minute or so photographers were allowed in at the top of the secret meeting).
With sarcasm dripping from his voice, Sen. Josh Hawley (R-MO) told Raw Story: “All those forums, they were critical. That really built momentum. It's important to have those cocktail parties with all of the biggest companies in the world, because they always want us to take action.
“Those were a total joke,” Hawley added, all but rolling his eyes.
Pre-Chat GPT, Hawley was one of the loudest voices on Capitol Hill calling to overhaul the internet. He wanted to end the blanket protection from prosecutions tech companies are afforded by Section 230 of the Communications Decency Act of 1996, which shields websites from the content others publish in them.
Now, Hawley says the threat of lawsuits would force Silicon Valley’s finest to fall in line and finally police their content for, say, child porn and AI-driven deepfakes alike.
“We need a change in legislation to make it clear that AI companies cannot hide behind Section 230,” Hawley said.
Hawley says the Biden deeepfake in New Hampshire is nothing compared to what’s coming ahead of elections in November.
“The voice is bad, but the videos are going to be really bad because they're at a point … I mean, you can't tell the difference. A lot of them now they're gonna get better and better and better, more and more quickly, and when you add the voice with the video, I mean, it's gonna be impossible for people not to know,” Hawley said.
So why hasn’t Congress acted?
“The companies,” Hawley said. “What’s going to have to happen is public outrage.”
And those tech companies surely have allies in Congress, including the relatively powerful ranking member of the Senate Commerce Committee, Sen. Ted Cruz (R-TX).
ALSO READ: Deadline demolished: Illinois congressman violates federal financial disclosure law
While many AI innovators — such as Chat GPT founder Sam Altman, to name one of many — have called for smart, proactive AI regulations, Cruz and other Republicans reject calls for new regulations for these new, ever-evolving technologies.
“If the Democrats push through restrictions on innovation and AI, it would be disastrous for America,” Cruz told Raw Story last year after leaving one of Schumer’s AI briefings.
While Hawley and others on the Senate Judiciary Committee have focused on the broader battle to unwind Silicon Valley’s current litigation shield, Sen. Amy Klobuchar (D-MN) is the main sponsor of an effort aimed specifically at combating deepfakes.
“I guess it's gotten front page interest now,” Klobuchar told Raw Story. “This is what I've said — hair on fire. We can't wait until next year to get this done. This is just going to keep happening.”
Klobuchar’s measure is the Protect Elections from Deceptive AI Act, which would outlaw fake AI generated content.
While state officials hope to prosecute the presidential spoofer in New Hampshire, Klobuchar argued that if her measure was law, the Department of Justice would also have a role to play in the investigation.
“It would have been banned,” Klobuchar said. “So at least going into it, people that might think that they could participate in something like that would know that it was against the law from the very beginning.”
While Congress has dithered, Klobuchar says some states are being proactive.
“They've done it in states blue and red. They've done it in Texas like that, but those things only apply to state campaigns,” Klobuchar said, adding that she hopes it can be taken up alone or as a part of a broader — and yet to crafted or introduced — AI measure. “That’s why this has got to be a priority in our AI legislation is doing it on its own or doing it as part of a package.”
Much of the focus in the media, and even among lawmakers, has been on watermarking AI generated content — think of it as a permanent digital stamp so anyone can trace the origins of suspect online material.
But Klobuchar says that only goes so far.
“That is not going to be enough, you're not going to have a fake Joe Biden make a call — or a fake Donald Trump — and then at the end, you go, ‘Oh, by the way, this was created by AI.’ It's just, it's not gonna work,” Klobuchar said. “So that's why you have to ban the actual deepfakes.”
Klobuchar says there’s no need to reinvent the wheel, either.
“TV does this all the time. TV decides there's ads they can allow, they look at them and say ‘do they meet the FCC standards or not?’” Klobuchar said. “Well, this would be the same kind of thing for deepfakes.”
Thing is, senators were warned — if behind closed doors — about the potential for a tidal wave of deepfakes.
“We could all predict that this was going to happen, so I hope that this will be a bipartisan effort to make sure that people are not lied to in this way. It’s terrible,” Sen. Mazie Hirono (D-HI) told Raw Story. “Since it's already upon us, we should do something about it to prevent this kind of abuse.”
While being in the midst of the 2024 election makes naysayers dismiss the efforts chances of passing this year, Hirono says the opposite should be true.
“I should think that there's more of an imperative to do something about it,” Hirono said.
