House passes bill to force transparency about use of AI in campaign materials

The Massachusetts House passed a measure Thursday that would require political candidates, super PACs, and others to disclose whether they use artificial intelligence-generated material in campaign ads or other materials, marking one of the Legislature’s first efforts to target “deepfakes” on the campaign trail.

House leaders tucked the Republican-filed language into a sweeping amendment they passed late Thursday as part of their ongoing voting over the chamber’s state budget proposal.

The measure comes in direct response to the rise of use of AI in campaign materials. In January, for example, a fake robocall imitating the voice of President Biden roiled Democrats ahead of the New Hampshire presidential primary, sparking a criminal investigation.

The proposal targets any audio or video “communications” paid for by a candidate, political parties, political action committees, or other groups that is “intended to influence voting for or against a candidate or ballot proposition in an election or primary.”

Any communication that uses so-called synthetic media, or content that’s “substantially produced” by AI, would have to include the message “Contains content generated by AI” at the beginning, the end, and during the video. Violators would face fines up to $1,000.

House Minority Leader Brad Jones, who filed the proposal, said there could be reasonable uses of artificial intelligence in campaigns, but he believes lawmakers need to start considering potential guardrails on how it is deployed.

“We gotta start having this conversation because we’re going to be chasing our tail if we don’t start thinking about it now, because we’re already here,” the North Reading Republican said.

The language the House passed may not be “perfect,” but it could be a start, Jones said. “It was more to kind of get the juices flowing,” he said.

The House was expected to wrap up voting on its state budget proposal as early as Friday, after which the bill would move to the state Senate.

Massachusetts would not be the first to regulate the use of AI in campaign ads. Washington state passed a law last year targeting so-called deepfakes by requiring that any election communication that contains synthetic media carry a disclosure, including for audio media. A candidate whose voice or likeness is used without the disclosure can also sue to block the ad under the law.

The fake robocall used in New Hampshire set off alarm bells around the country, including in Washington state. Steve Hobbs, Washington’s secretary of state, said in a statement at the time that the situation was “just the tip of the iceberg for 2024” and warned that it would likely only become harder to spot a deepfake.

“These false messages will get more polished and harder to tell from real ones,” Hobbs said.