Lawmakers are considering a bill that would ban the use of “deepfake” video technology in political advertising in Maine.

Supporters say the measure is needed because the technology can be used to manipulate and mislead voters, producing what one legislator called a “disastrous effect” on U.S. society and democracy. But critics countered that the bill is a threat to free speech.

Deepfake videos are built with digitized images, using a form of artificial intelligence known as deep learning. The technology, which is increasingly available on the internet, can be used to produce videos, still images or audios that are highly convincing and often impossible to identify as fake without the use of detection software.

In politics, the technology has been used to fake what elected officials have said or portray them in embarrassing, but false, situations.

The use of deepfake videos has been a growing concern for federal policymakers. Prominent politicians have been smeared by its use, especially on social media sites where the videos can be broadly shared. U.S. Sen. Angus King, an independent from Maine, has also voiced concern over the technology’s use in politics.

“We are lucky to live in a society with technological advancements that past generations could have never imagined,” Sen. Rebecca Millett, a Democrat from Cape Elizabeth who sponsored the bill, told the Legislature’s Veterans and Legal Affairs Committee on Wednesday. But those advancements, she said, can also be used to undermine the political process and public trust.

“It’s not hard to imagine how destructive this technology could be in an election,” Millett said.

Millet’s bill, L.D. 1988, would prohibit publication and distribution of deepfake media of a candidate within 60 days of an election. The bill would allow a targeted candidate to seek a court order to block publication and to pursue civil action against whomever published or distributed the material.

The ACLU of Maine and the Maine Heritage Policy Center, a conservative research and public policy group, submitted written testimony opposing the legislation. But the bill did gain the support of the League of Women Voters of Maine.

Ann Luther, a Trenton resident and volunteer league member, said having a well-informed electorate is key to the organization’s mission, and she offered qualified support for the measure.

“This bill will not obviate the need for each of us to be alert consumers of the news,” she said. “We all need to apply healthy skepticism to click-bait headlines and determine the source and veracity of the information we consume – but it can help block the worst examples of malign misinformation.”

Luther also noted, however, that the measure “does raise famously sticky constitutional issues of allowing the government to scrutinize the content of speech …”

Republican committee members peppered Millett with questions concerning free speech rights.

Rep. John Andrews, R-Paris, quizzed Millett on several aspects of the bill, and asked how the American Civil Liberties Union responded to a similar bill passed in California that Millett relied on as a model for her legislation.

Millett said she didn’t believe the ACLU had weighed in on the California measure, but Andrews read aloud a quote he attributed to Kevin Baker of the California ACLU.

“Despite the author’s good intention this bill will not solve the problem of deceptive political videos, it will only result in voter confusion, malicious litigation and suppression of free speech,” Andrews read. “So, I don’t think they supported it.”

The Maine Association of Broadcasters offered neutral testimony, but urged the committee to retain provisions that would protect the use of deepfake material for illustrative purposes in news reporting.

Suzanne Goucher, the association’s president, said Millett’s bill would have little impact in the social media realm. She pointed out that social media platforms do not undergo the federal scrutiny that broadcasters do with political advertising.

“Facebook, Twitter, the social media, that’s the Wild West,” she said.

Both U.S. Speaker of the House Nancy Pelosi and former President Barack Obama have been the subjects of deepfake videos in recent years. The altered video of Pelosi, which was slowed down to make it appear that she was slurring her words, was dismissed by experts a “cheap fake” because it didn’t use the more sophisticated technology.

However, the video was viewed more than 3 million times on social media, demonstrating the potential reach of intentionally misleading media.

The U.S. House Ethics Committee has taken up the issue of doctored material, and on Tuesday issued a memo warning lawmakers that sharing such images risks violating House ethics rules, the Associated Press reported.

On another election-related issue, lawmakers took testimony Wednesday on a bill that would prohibit Clean Election candidates from taking a job with any organization or business paid more than $10,000 for services to their campaigns.

The change was offered by Rep. Kent Ackley, an independent from Monmouth, who learned that former state Sen. Garrettt Mason, R-Lisbon, had paid a consulting firm more than $100,000 from his Clean Election campaign funds during his failed bid for governor in 2018. Mason later went to work for the firm.

 

« Previous

Social Security administrators say their agency is getting barraged with scam complaints

Next »

Ukraine ‘quid pro quo’ not an impeachable offense, Trump’s defense says