The Daily Gazette (Op-Ed): State must protect voters from AI fraud

There is no doubt among political scientists and technologists alike that generative Artificial Intelligence (AI) threatens democracy and election security.

Disinformation more generally has become increasingly prevalent in the last decade, but AI brings this to an entirely new level.

As advocates for government accountability, we have seen firsthand how legacy media and journalism have struggled to keep pace with the ever-more volatile environment of disinformation on social media.

It’s critical that New York legislators account for the harms that AI can cause in elections during this session ahead of the November presidential election.

Thoughtless misinformation and deceptive disinformation have been on the rise and have stoked the flames of the violent political climate we have today. AI electioneering is now poised to douse the flames in fuel.

Not too long ago, just next door in New Hampshire, we saw how deepfakes have the potential to confuse or mislead voters to impact elections.

Synthetic audio using AI technology to resemble President Biden’s voice urged voters not to vote in our recent primary.

Specifically, the audio stated, “It’s important that you save your vote for the November election… Voting this Tuesday only enables the Republicans in their quest to elect Donald Trump again.”

Data shows that 76% of the synthetic audio targeted New Hampshire, 12% targeted Boston, and 12% targeted other areas, reaching up to 25,000 people.

However, this wasn’t the first time deceptive AI-generated content reached voters.

AI deepfakes have been increasingly used in media and campaign communications to spread disinformation about candidates.

For example, in 2023, Florida Gov. Ron DeSantis’s campaign released a deepfake image of former President Trump hugging and kissing Dr. Anthony Fauci on the nose.

This tactic from the DeSantis campaign reveals how AI tools can easily be used to deceive voters during an election. It also shows the alarming reality of how AI can affect the 2024 presidential race.

We need common sense reforms and regulations to address the many gaps in the arcane laws comprising our current election policy.

This is an issue that is affecting candidates and politicians across the political spectrum, and the noise needs to be channeled into action and coupled with the bipartisan legislation that we need.

As the times evolve, so must our laws. Luckily, legislation is moving around the nation that addresses these new issues.

Statehouses across the country are passing bills that would ban AI deepfakes in election communications.

Here in New York, Assembly bill A7106 and Senate bill S8631 would require that any deepfake of a candidate or political material made using AI has a disclosure stating that it was made with AI.

These are important pieces of legislation that would bring our law up to date and protect our elections from fraud and disinformation as we head into the new frontier of AI.

With the New York Legislature currently negotiating its budget for the year, it has the power to address this now by including wording in the budget to address the abuse of AI in elections.

Including this in the budget will solidify New York’s stance against fraudulent AI electioneering and symbolize a commitment to protecting our elections from the threat of weaponized AI.

We know that we can address the threat of AI-generated disinformation with common sense reforms.

But, at the same time, we cannot rest on our laurels and expect the government to work everything out.

We need to come together as informed and concerned citizens of this state to show our legislators that we want them to act now to ensure that this technology doesn’t negatively impact our democracy and elections.

Susan Lerner is the executive director of Common Cause and a founder of Let NY Vote. Jonah Minkoff-Zern is the co-director of Public Citizen’s Democracy Campaign.

 

To view this piece online, click here.