Their concerns reflect bigger questions facing candidates about how to adequately prepare for a volatile technology landscape that lacks regulation as the high-stakes election year ramps up into full gear.
“AI has tremendous potential, but without the appropriate guardrails, it also poses grave dangers,” said Representative Lori Trahan of Westford, in a statement to the Globe. “We believe it’s vitally important that the [Federal Election Commission] issues clear guidance without delay … so bad actors can be held accountable for using AI to deceive voters.”
AI is already playing a growing role in the political ecosystem, with generative tools like ChatGPT making an appearance in campaign ads and on social media platforms. Candidates can also harness the technology as a tool to target specific audiences with precise messaging.
An ad by Never Back Down, a super PAC that supported Florida Governor Ron DeSantis in his presidential campaign, used artificial intelligence to mimic Trump’s voice criticizing Iowa Governor Kim Reynolds. Trump co-campaign manager Chris LaCivita told Politico it was a “desperate attempt … to deceive the American public.”
In another instance, the Republican National Committee ran an ad featuring AI-generated dystopian images of what they said the country could look like under a second Biden term. The next month, Democrats introduced legislation that would require disclosures of AI-generated images or video in political ads, though neither chamber of Congress has voted on it.
In New Hampshire, unauthorized robocalls ahead of the presidential primary purporting to be a message from Biden urging voters not to vote in the primary election on Jan. 23 sparked a state criminal investigation. A New Orleans magician told NBC News that a consultant for a rival campaign had hired him to create the audio.
Those robocalls were why Kathy Sullivan, a Biden supporter and major Democratic player in New Hampshire politics, suddenly found herself an expert on AI use in campaigns. The calls were erroneously linked to her number.
“Every candidate needs to have [AI] on their radar, because when it comes to election interference, it’s not something that just happens at the top of tickets,” Sullivan said. “If you’re in a congressional or senate race, you should be totally geared up to be prepared. What’s your plan going to be if this happens to you? Who are you going to talk to? How are you going to get the word out?”
Sullivan’s advice for candidates includes asking their supporters to keep an eye out for information that seems suspicious or out of character for their campaigns, and to have dedicated staff looking out for this type of threat.
Representative Ayanna Pressley of Boston noted that there’s been a rise in political threats and violence in recent years, but at least for her, AI doesn’t feel like an immediate concern. She said she was not aware of any instances of her team having to respond to the misuse of AI related to her, and had not given her team instructions to take additional preparations going into this year’s elections.
“I’m sure there’s already some sort of plan in place,” Pressley said. “But it’s not anything that I’ve prompted.” Her campaign said in a statement that the threat of AI only highlights the importance of community organizing.
Artificial intelligence experts worry that the technology will aid foreign interference in the elections, making it easy for bad actors to inundate voters with conflicting information and sow confusion.
“My biggest concern is people being so overwhelmed and finding it so hard to get information that they just throw their hands up. And that’s really where foreign adversaries win, when voters decide they’re going to disengage and not participate in our democracy,” said Jiore Craig, a resident senior fellow at the Institute for Strategic Dialogue.
And while many experts and political observers said it’s too soon to tell how big of a factor artificial intelligence — and its misuse — will play in this year’s elections, there’s a widespread belief that the onus shouldn’t be on voters to have to figure out whether content related to elections is real.
“Artificial intelligence could be an example of where we’re just not thinking quickly enough,” said Representative Seth Moulton of Salem. “Where’s the party’s task force on AI? Where is the world-leading effort that the Democratic party should be leading to show how it can be used? Why is there no AI policy from the Democratic party?”
The Democratic Congressional Campaign Committee, House Democrats’ campaign arm, hasn’t sent its candidates uniform guidance or provided briefings on AI. But DCCC spokesperson José Muñoz said the group is providing online monitoring on behalf of Democratic House candidates, especially those in battleground districts, and urging those candidates to provide voters with ample information about what they stand for to get ahead of misinformation.
“The focus really has been about being able to provide that monitoring and identifying potential threats,” Muñoz said.
Not everyone views AI skeptically— for some, the potential benefit outweighs the risk of malicious use.
“The purpose of a campaign is to allocate limited resources as efficiently as possible and to deploy a message from a candidate that gives you the best chance to win a majority,” said Zac Moffatt, Republican strategist and former digital director for Mitt Romney’s 2012 presidential campaign. “To me, all of the tools of AI and everything else … helps you identify the people who are most open to your message, helps you identify places where you have weaknesses.”
Lawmakers like Auchincloss and Senator Ed Markey — the only member of the Massachusetts delegation not up for reelection this year — have proposed legislative ideas for regulating AI. But many said it’s not yet clear how big of a role it will play in elections.
“Three words: to be continued. Because I think we’re going to see things that we didn’t even anticipate,” said Representative Bill Keating of Bourne. “That’s what’s fearful.”
Charlotte Ehrlich can be reached at charlotte.ehrlich.globe.com. Lissandra Villa de Petrzelka can be reached at lissandra.villa@globe.com. Follow her @LissandraVilla.
Source: There’s a rise in the use of A.I. in campaigning— are Mass. congressional candidates
