MONTPELIER, Vt. — Vermont has enacted a law requiring disclosure of certain uses of artificial intelligence in election-related communications, joining a growing number of U.S. states seeking to address concerns about AI-generated political content ahead of upcoming elections. The measure, known as S.23, was signed into law by Governor Phil Scott on March 5, according to state legislative and advocacy records.
The law requires campaign media that uses AI-generated images, audio, or video within 90 days of an election to include a clear disclosure informing viewers that the content was created or manipulated using digital technology or artificial intelligence. The legislation also establishes rules governing deceptive synthetic media intended to influence voters.
Supporters of the measure said it is designed to increase transparency as increasingly sophisticated AI tools make it more difficult for voters to distinguish authentic content from manipulated material. During legislative debate earlier this year, lawmakers said the proposal was intended to provide voters with information rather than prohibit the use of AI technology in political advertising.
“It’s extremely hard to tell what’s real and what isn’t,” state Senator Ruth Hardy, the bill’s lead sponsor, said during consideration of the legislation. “We need to make sure that especially in the months leading up to an election that voters are aware” when content has been manipulated, according to Vermont Public.
The law focuses on synthetic media that could mislead voters by depicting speech, actions, or events that did not occur. According to summaries of the legislation, disclosures must be visible and understandable to audiences viewing AI-generated campaign content. Enforcement provisions are included under state election and consumer protection authorities.
Vermont lawmakers considered the measure as states across the country evaluate how to regulate political uses of generative AI. Legislative testimony reviewed during committee proceedings noted that several states have adopted disclosure requirements or other restrictions aimed at addressing AI-generated election content and so-called deepfakes.
Supporters, including the Vermont Public Interest Research Group, said the law would help voters identify AI-generated political material. Critics in other states have raised concerns that similar measures could face legal challenges related to free speech, though details regarding any potential challenges to Vermont’s law remain unclear.
The law is now in effect and is expected to apply during Vermont’s 2026 election cycle, according to legislative and advocacy records.
