Wyoming lawmakers grapple with challenges of AI regulation, debate best way to approach it

May 21—CHEYENNE — State legislators around the country are tackling the issue of regulating deepfakes, an enormous feat to take on as the law rushes to catch up with the growth of advanced artificial intelligence technology.

Deepfakes, or digitally altered content generated with AI, has national security officials concerned. The U.S. Department of Homeland Security (DHS), National Security Agency and the FBI have all recently delivered warnings against the rapidly advancing technology.

"Deepfakes and the misuse of synthetic content pose a clear, present, and evolving threat to the public across national security, law enforcement, financial, and societal domains," the DHS report stated.

An NPR article published in February reported that New Hampshire voters received a robocall impersonating President Joe Biden that tried to persuade voters not to vote in the upcoming Democratic primary.

Wyoming lawmakers discussed a bill on Monday that was sponsored by the Legislature's Select Committee on Blockchain, Financial Technology and Digital Innovation Technology in the recent 2024 budget session.

Senate File 51, "Unlawful dissemination of misleading synthetic media," would have prohibited the distribution of synthetic media, or deepfakes, with the intent to deliberately mislead people and spread misinformation. The bill required a disclaimer be posted with any digital content that was altered with AI technology.

SF 51 made it halfway through the session before it died in a House committee, where some lawmakers argued the bill impeded on an individual's freedom of expression.

"It's a real tension. It's something that is always a challenge when you are providing limitations on speech," Sen. Chris Rothfuss, D-Laramie, told the Wyoming Tribune Eagle on Thursday. "How do you draw those lines? Where do you draw those lines? And how do you make sure that you have a compelling state interest?"

Narrowing it down

Sen. Affie Ellis, R-Cheyenne, said on Monday her biggest concern is with the bill's broad language that could lead to unintended consequences. She would've been more comfortable with the bill, she said, if it were narrowly tailored to a certain area, such as election campaigns.

"If it were just tailored to campaigns ... then we can have discussions about libel, slander," Ellis said. "They fit better separated out, rather than as one big bill, because that's where I start getting lost and seeing unintended consequences."

Practicing attorney Matthew Kaufman, who previously served on the Legislature's Blockchain Taskforce in 2019, agreed with Ellis, adding that the broad language would render the statute useless. The "intent to mislead" form of tort in the bill, he said, is a "unique circumstance."

"I'm worried that we're creating a standard that's going to be very difficult to meet," Kaufman said.

This is also an issue that lawmakers at both the state and federal are seeing the development of AI and related governance laws "play out in real time," Kaufman said. A lot of this has yet to be dictated at the federal level.

"I don't want to see us waste too much time on those things that we just don't have control over," Kaufman said.

Rothfuss previously told the WTE he was confident in the bill as it was written. Current state and federal laws don't provide a remedy for people who have been misled, he said at the meeting, and that's what this bill was trying to do.

"Do we individuals have a right to know if something's true? As of right now, you have no such right," Rothfuss said.

If a Wyomingite were misled by an out-of-state political candidate, such as a presidential candidate for example, Rothfuss questioned if there's anything in current state law that would protect the voter.

"In my view, there is a value to knowing whether you've been handed the truth or not," Rothfuss said. "But there's no mechanism in any statute to provide surety."

Unintended consequences

Members of the committee considered the different uses of AI technology that, while possibly misleading, are considered perfectly reasonable. Rep. Daniel Singh, R-Cheyenne, brought up the fact that AI-generated content is a common practice in advertisements.

Legislative Service Office staff attorney David Hopkinson said the bill specified that information is considered misleading if a "reasonable person" is unable to differentiate fact from fiction. Hence, any AI-generated content in an advertisement, TV show or film would be perceived by a reasonable person as fiction.

Kaufman said part of the problem is that the bill is broadly applied to any circumstance. This wide applicability of the statute makes it difficult to enforce or for a person to provide proof of injury.

"I'm thinking ahead of unintended consequences," Kaufman said. "They have to 'knowingly and intentionally' disseminate with an intention to mislead. That's a lot, right there, to (prove)."

This broad terminology could end up as "unwieldy and unuseful" in a court of law, unless it were narrowly tailored, as Ellis suggested, to election campaign laws, he said. Certified public accountant David Pope said he saw "a lot of value in taking smaller bites."

"If we have to confine it to political actions in the near term, what it will allow us to do is work through the language, work through the concepts and build a framework that I would love to see expanded in the future," Pope said.

Committee members decided to bring back discussion on SF 51 at their next meeting, which will be held July 1-2 in Sheridan.

Hannah Shields is the Wyoming Tribune Eagle's state government reporter. She can be reached at 307-633-3167 or hshields@wyomingnews.com. You can follow her on X @happyfeet004.