How Federal Agencies Are Preparing for AI-Driven Election Threats in 2024

Federal election security officials are managing several overlapping challenges heading into the 2024 election cycle. In May 2024, DHS Secretary Alejandro Mayorkas identified three main threat areas: cybersecurity vulnerabilities, physical risks to election officials and poll workers, and disinformation campaigns designed to mislead voters. The Department of Homeland Security called the current threat landscape an "unprecedented array of election threats," pointing to how technological advances and ongoing geopolitical tensions are colliding at the ballot box.

AI-Generated Content Is Already Here

In February 2024, U.S. news outlets reported what appears to be the first documented case of AI-generated audio targeting a U.S. election—specifically, synthetic audio mimicking a presidential candidate's voice. This wasn't a theoretical warning; it happened.

The significance lies in what it reveals about technical barriers. Two years ago, voice cloning required specialized equipment and large amounts of training data. Today, consumer-grade software and relatively small audio samples can accomplish the same thing. For election security teams, this shift changes the entire threat model, especially in the final weeks before voting when responding quickly to false content becomes critical.

A major technology company assessed in late 2023 that the 2024 presidential race could mark the first time multiple hostile foreign governments attempt to interfere with an American election simultaneously. Combined with the emergence of actual AI-generated content targeting candidates, this suggests election security professionals are facing both more capable adversaries and more of them than in previous cycles.

The broader context here is worth acknowledging. We have seen this pattern before, when new communication technologies intersected with electoral processes—the shift from print to radio, radio to television, television to social media each created new vectors for both legitimate campaigning and malicious interference. The current AI-enabled disinformation landscape follows that same trajectory, though the pace has accelerated considerably. Where previous election interference operations required significant resources and coordination, AI tools can now produce sophisticated content quickly and cheaply, potentially by small teams with modest budgets.

Federal Monitoring Expands

The Justice Department announced plans to monitor compliance with federal voting rights laws across 86 jurisdictions in 27 states for the November 5 election. Federal observers will oversee procedures at polling places and ballot-counting facilities in eligible jurisdictions, including polling sites in California and New Jersey, among others.

The Civil Rights Division's election monitoring program focuses on ensuring transparency, ballot security, and compliance with federal law. Justice Department officials have characterized this monitoring as routine compliance work rather than a response to specific threat intelligence. The scope—86 jurisdictions across 27 states—represents one of the larger federal election oversight deployments in recent cycles.

Immigration Enforcement and Polling Site Concerns

A separate concern has emerged around the potential presence of Immigration and Customs Enforcement agents near polling locations. Federal law explicitly prohibits federal officers from interfering in elections and bars armed federal agents from deployment at election sites, including polling places, vote-counting facilities, and ballot drop boxes.

Despite these legal protections, a group of state secretaries of state demanded that Trump's DHS nominee pledge not to deploy ICE agents to polling sites during elections. Their concern appears to center on the risk of voter intimidation, particularly in communities with significant undocumented populations.

DHS officials have provided assurances through multiple channels. A Department official told state election administrators that immigration agents would not be stationed at polls during elections. The White House stated that President Trump has not discussed formal plans to deploy ICE agents to polling sites, and DHS confirmed it was not planning immigration operations targeting polling locations.

The underlying tension reflects a practical complication in how federal law enforcement operates during elections. Immigration enforcement and election security fall under different legal authorities, but when they intersect at the community level, they create real complications for election administrators trying to encourage broad voter participation.

What Comes Next

Looking ahead, the technical fundamentals suggest AI-enabled disinformation will be an escalating challenge rather than a temporary one. Generative AI capabilities continue to improve rapidly while the cost of computing power declines. The barrier to entry for sophisticated disinformation operations will likely continue falling, making detection and rapid response increasingly critical for election security teams.

The broader operational environment—spanning cybersecurity threats, physical security concerns for election workers, AI-generated content, and immigration enforcement coordination—presents a more complex challenge set than election security teams have previously managed. The federal response appears to acknowledge this complexity through expanded monitoring operations and inter-agency coordination. The real test will come during the election cycle itself, when the scale and sophistication of actual interference attempts becomes clear.