Bipartisan House bill aims to prepare election workers for AI

Ahead of a Senate markup this week on a trio of bills governing artificial intelligence in elections, a bipartisan group of House lawmakers on Monday introduced companion legislation aimed at preparing election administrators for challenges posed by the technology.

The Preparing Election Administrators for AI Act from Reps. Chrissy Houlahan, D-Pa., Brian Fitzpatrick, R-Pa., Abigail Spanberger, D-Va., and Doug Lamborn, R-Colo., would require the Election Assistance Commission to team with the National Institute of Standards and Technology on a report that delivers voluntary guidelines for election administrators on the related risks and benefits of AI. 

Houlahan said in a press release that the legislation — the House counterpart to a March bill from Sens. Amy Klobuchar, D-Minn., and Susan Collins, R-Maine — aims to “strengthen trust and confidence” in the electoral system, something she called “a shared responsibility that transcends party lines.”

“With the rise of artificial intelligence, we must equip our election administrators with the necessary tools and guidelines to safeguard our democratic process,” Houlahan said in a press release. “This bipartisan bill underscores our commitment to fair and secure elections by addressing the potential risks posed by AI, including misinformation dissemination and cybersecurity threats.” 

Given the potential for AI-fueled trickery in the administration of elections, Fitzpatrick said “it is essential that state and local officials are prepared and provided with the resources to keep the election process secure,” adding that Congress must keep working to “ensure the public trust and the integrity of our elections are not compromised.”

The bill calls on the EAC and NIST to deliver their report to Congress and state and local election offices within 60 days of its enactment. The report would also detail the cybersecurity risks of AI systems in elections.

There are also callouts in the bill for the EAC and NIST to outline how election offices should respond to information generated and distributed by AI that compromises the sharing of accurate election information. The report would also detail how AI could impact the spread of election-related disinformation, undermining public confidence in the process.

Spanberger said in the press release that the legislation is important for not only strengthening election infrastructure but also to “raise our defenses against bad actors,” citing Russia and China specifically. 

“Americans need to have confidence in their vote — and our hardworking election officials need the tools and training necessary to navigate AI-related vulnerabilities in our election systems,” she said. 

Officials in Arizona, Minnesota and other states across the country have created training exercises for local election workers to identify and combat AI-generated content, and voting this November has been dubbed by some to be “the AI election.” At last week’s RSA Conference in San Francisco, however, federal officials told reporters that physical violence against election workers represents a bigger threat than AI. 

In addition to marking up the Preparing Election Administrators for AI Act, the Senate Rules Committee on Wednesday will also consider the chamber’s Protect Elections from Deceptive AI Act and the AI Transparency in Elections Act of 2024. House Reps. Derek Kilmer, D-Wash., and Tony Gonzales, R-Texas, on Tuesday introduced a companion bill for the former, which targets AI-generated content in political campaign advertising.

This story was updated May 14, 2024, with information on the introduction of the House’s Protect Elections from Deceptive AI Act.