Vermont Considers Criminal Penalties for AI Election Misinformation

Vermont lawmakers are deliberating on legislation that would criminalize the distribution of AI-generated election misinformation in the period leading up to Election Day, positioning the state to join a growing national effort to regulate artificial intelligence in political discourse.

The bill, sponsored by Republican Representative Woodman Page of Newport, would make it a criminal offense to post AI-generated disinformation about elections or advertise such content during campaign seasons. Additionally, it would mandate high-traffic online platforms to properly label synthetic, inauthentic, or false content.

“Vermont is behind the times when compared to larger states,” Page said, noting that the legislation aims to align Vermont with standards being established across the country. States including California and New York introduced similar regulatory frameworks in recent months.

Page emphasized that the bill responds to rapidly evolving AI capabilities that pose emerging threats to electoral integrity. “It’s unfortunate that we need such a bill,” he remarked, pointing to growing concerns about AI’s potential misuse in both politics and private affairs.

The proposed legislation has garnered support from political science experts like VTSU Castleton professor Rich Clark, who views it as necessary to combat election misinformation.

“We need to get our hands on misinformation, as it is really detrimental to our elections,” Clark said. He explained that AI technology exponentially increases the volume of false information circulating online, making it increasingly difficult for voters to distinguish fact from fiction.

Clark referenced a recent incident involving altered images of a Minneapolis church protester that appeared in White House media. The original photo showed the protester with a vigilant expression, but the manipulated version depicted the individual sobbing with a darker skin tone – demonstrating how easily images can be altered to change public perception.

“I’ve often said it’s almost more important that voters believe elections are fair than elections actually being fair because if elections are fair and nobody believes it’s fair, what does it matter?” Clark noted, highlighting the critical relationship between perceived and actual electoral integrity.

The bill enters uncharted constitutional territory, however, raising questions about First Amendment protections for political speech. The legal status of AI-generated content remains ambiguous, with some arguing that such material doesn’t qualify as human speech protected under the Constitution.

VTSU Castleton political science student Quin Forchion acknowledged this complexity. “In a different way that we haven’t seen before, it isn’t completely our independent speech when it comes to AI,” Forchion said, describing the blurry distinction between human expression and AI-altered communication.

Forchion suggested that AI tools aren’t neutral instruments but often reflect the biases and agendas of their creators. He pointed to instances where AI systems have demonstrably been influenced by changes to their underlying code, citing the controversy surrounding Grok, the AI used on X (formerly Twitter), which reportedly began generating antisemitic content after code modifications allegedly favored conservative viewpoints.

The exact prevalence of AI-generated content on social platforms remains disputed, with estimates ranging dramatically from as low as 2% to as high as 71% of posts on some platforms, according to various sources.

Both Clark and Forchion emphasized the fundamental importance of safeguarding electoral systems from deliberate misinformation. “It needs to be treated as information being sacred,” Forchion stated, underscoring the bill’s goal of preserving democratic integrity in an era of rapidly advancing technological capabilities.

As Vermont lawmakers deliberate this legislation, they join a nationwide conversation about balancing free speech protections with the need to prevent technological tools from undermining democratic processes.