New York enacts law to curb AI deception in elections

May 31, 2024 Robert Abruzzese, Courthouse Editor
Share this:

In a move meant to safeguard the integrity of its elections, New York has passed a law targeting the use of artificial intelligence (AI) in political communications. The legislation, adopted in mid-April 2024, addresses the growing concerns over AI’s potential to manipulate voter perceptions through deceptive practices.

The impetus for this legislative action stemmed from a recent incident in early 2024 when an audio clip that appeared to feature a prominent political party chairman using profanities and disparaging an incumbent member of Congress went viral in New York City. The clip, which seemed genuine, was later revealed to be an AI-generated deepfake, highlighting the urgent need for regulatory measures.

Gov. Kathy Hochul, who championed the legislation, discussed the importance of setting boundaries for the use of advanced technology in political campaigns when she made the announcement on the change in the law.

“This legislation will help to set important guardrails around the use of this cutting-edge technology and give law enforcement the tools it needs to go after bad actors, and I will work closely with the Legislature to make sure we get this done,” Hochul said. 

The new law, part of the Fiscal Year 2025 state budget, amends Election Law 14-106 to impose stringent requirements on political communications altered by AI. Any entity distributing or publishing such communications must disclose if the material has been manipulated using AI technology, provided it could reasonably be perceived as authentic. This disclosure must be clear and conspicuous, ensuring that voters are not misled by AI-generated content.

Specifically, the law mandates that any political communication altered with AI include a disclaimer stating it has been manipulated. This applies to all forms of campaign materials, including advertisements, pamphlets, flyers, brochures and digital content. For audio communications, the disclaimer must be clearly spoken at the beginning and end of the recording and at two-minute intervals for longer segments.

Exemptions to the law include media considered satire or parody, bona fide news reporting with the required disclosure, and content distributed by platforms that did not create it but made a good faith effort to verify its authenticity.

The New York State Board of Elections is expected to propose regulations and provide further guidance on the law’s implementation, aiming to ensure compliance and protect the electoral process from AI-driven deception.

Proving the use of AI in political communications poses significant challenges. Many programs designed to detect AI-generated content are not foolproof, give false positives, and often struggle to keep pace with the rapid advancements in AI technology. These detection tools can sometimes fail to distinguish between authentic and manipulated media, making enforcement of the new law complex.


Leave a Comment


Leave a Comment