Congressional campaign using AI to help voters learn about their candidate
PALO ALTO — Artificial intelligence is being put to use in elections across the United States this year. And experts in the field have been warning about AI's role in the electoral process as deep fakes have been caught spreading false messaging to voters.
But while the software has yet to see firm regulation, one campaign in Silicon Valley is leaning into its use to set an example for campaigns to come.
The Dixon campaign for the district 16 Congressional seat held by Anna Eshoo turned to AI to get out messaging about candidate Peter Dixon.
They're not shy about making its use known. Afterall, he's running in a district that is considered to be the birthplace of AI.
Jennifer is the name of the artificial intelligence volunteer that makes calls to voters that otherwise would be done by dozens of human volunteers that may be hard to find in a short time to launch and execute a campaign.
Dixon said the idea was to reach as many voters as possible in the three-month campaign cycle.
"It allows voters to spend as much time as they would like to engage with the capability and getting a sense truly who I am and what I stand for so they can be informed when they go to vote," Dixon told CBS News Bay Area.
His is one of a small handful of campaigns across the country deploying the new technology from political AI company Civox in the run up to the primary election.
"Jennifer's ability to answer questions accurately, not sort of guess if there wasn't a real answer and say 'I'm not familiar with Peter's position on that,' and really be a good steward answering questions honestly and truthfully," Dixon said of the software.
Jennifer isn't the only voice reaching voters. KPIX spoke with one version called Ashley who answered questions about economic, foreign and social policies of the Dixon campaign. The call was just under three minutes.
The AI caller gave detailed, factual, responses that outlined the candidate's platform and provided conversational replies including the response: "great question."
But the potential of AI's wider presence in the electoral process has prompted some state lawmakers to introduce legislation that would protect voters and candidates alike from deceptive content.
Drew Liebert is the director of the California Institute for Technology and Democracy — a cosponsor of five bills that focus on regulating AI in state elections. He said his top priority is preventing the seamless spread of misinformation.
"We've already seen it for example in New Hampshire where the alleged voice of the president of the United States was discouraging people saying, 'Don't vote in the primary; vote in the general.' And what that was about was making sure Democrats didn't vote in the election," Liebert told CBS News Bay Area.
The proposed legislation seeks to require social media companies to label AI generated content, provide verification tools for images possibly created by AI, and ban deceptive deep fakes for 120 days before and 60 days after an election.
"We're on the precipice of the first real AI election," said Liebert. "We recognize that our traditional way of protecting democracy are inadequate now. We all need to recalibrate how to protect our democracy."
The pioneering technology is not only being deployed in the CA district 16 race, but it is also in the California senate race to fill the seat held by the late Dianne Feinstein and races in Ohio and other key states.
While regulators don't appear to be jumping to build barriers with this AI campaign volunteer, tech giants are getting ahead of potential scrutiny by pledging to help prevent the spread of misinformation in the 2024 election.
In February, 20 companies including Amazon, Google, Meta, Microsoft, and Open A.I. signed the pledge.