AI and Elections: Curbing Deceptive Ads Without Stifling Speech

AI and Elections: Curbing Deceptive Ads Without Stifling Speech

An NCSL Town Hall covers efforts at the state and federal levels to address the growing influence of artificial intelligence in elections.

State Legislatures NewsPolicyArtificial IntelligenceElections

(cyano66/Getty)

States and the federal government are grappling with how best to manage the challenges of artificial intelligence when it comes to administering elections, according to NCSL policy experts who spoke during an NCSL Town Hall on AI and elections.

"So much of the conversation about AI right now is about how AI could cause harm in elections, says Adam Kuckuk of NCSL's Elections and Redistricting Program. "But just like most tools, AI is neutral. It can be used for good or foul. A hammer can be used to build a house, but it can also be used to knock somebody over the head."

Lawmakers are trying to understand how AI can aid elections or disrupt them, and the pace of AI's evolution means there's always something new to understand, Kuckuk says.

Related: NCSL's Elections 2024 Town Hall Series

This area is so new for lawmakers that, according to NCSL's recent report on state and federal actions, it goes by several different terms, including synthetic media, deceptive media, deepfake and a variety of others that encompass the misleading use of media to influence an election. And states don't even agree on a universal definition of AI.

So far, at least 19 states have enacted laws about using AI in elections. California and Texas started the trend in 2019. Most other state laws were enacted in the last two years, Kuckuk says, and many are so new they haven't been tested in court.

One big challenge for legislators is how to ward off AI that deceives voters without stepping on First Amendment rights, he says. Rather than prohibiting AI in political messaging, states generally have chosen to require that its use be disclosed, so messages must include a statement much like those declaring who funded the ad.

In addition to addressing the fast-changing abilities of AI, states must keep an eye on what federal laws will do to change the landscape.

Sanam Hooshidary, of NCSL's State-Federal Affairs Division, says numerous bills have been proposed in Congress, though none have passed so far.

"Federal lawmakers are predominantly concerned about setting standards when it comes to transparency and preventing any kind of manipulations, and even preventing foreign interference through AI-driven disinformation campaigns," Hooshidary says.

She notes the Cybersecurity and Infrastructure Security Agency works with the FBI and other federal agencies and with state and local election officials to detect problems and educate the public.

"CISA plays a major role when it comes to protecting elections against cybersecurity risks, and especially now when it's AI-driven threats like deepfakes and maybe other manipulated content that malicious actors are using and can use to weaken the trust in our electoral process," Hooshidary says.

For all the uncertainty about the future of AI, Kuckuk says one thing is sure: It will keep rapidly evolving.

"I'm sure we'll see many other different approaches to requiring digital watermarks or something that hasn't even been thought up yet," Kuckuk says. "The landscape of state legislation is changing very dramatically, and it might change even more dramatically if Congress ends up enacting legislation regulating (AI's) use."

Kelley Griffin is the host and producer of NCSL's "Across the Aisle" podcast.