The legislation, sponsored by Republican state Rep. Wendy Hausman of St. Peters, combines bills offered by at least 10 lawmakers. And while it covers a range of digital media, Hausman described the legislation during debate last week as an effort to protect Missourians from the dangers of AI.
“This bill draws a bright line,” Hausman said. “If you intentionally use AI to harm, deceive or exploit, there are consequences.”
It would be a felony to share or threaten to share an AI-generated or other digital depiction of someone to harass, threaten or harm them, with a maximum penalty of four years in prison. The maximum penalty would be 10 years if the image depicts a minor.
Social media platforms would be required to implement age verification measures to ensure youth under 16 can’t create social media accounts and allow parents to oversee social media use by 16- and 17-year-olds. The bill would prohibit social media platforms from targeting minors with paid advertising or “addictive or manipulative design features” like “infinite scroll” mechanisms and auto-playing content without time limits.
Additionally, individuals or companies that develop AI or make it available to the public would be barred from advertising their product as a mental health professional or capable of providing therapy or mental health diagnoses. They would face a penalty of $10,000 for violating this measure, or $20,000 for subsequent offenses.
The House voted 145 to 3 to advance the legislation to the Senate, with three Republicans in opposition.
The bill’s advancement comes as lawmakers grapple with the stakes of establishing AI guardrails after President Donald Trump issued an executive order in December warning states against passing “onerous” AI legislation and threatening to withhold federal funding for rural high-speed internet access from those that do.
The executive order established an AI Litigation Task Force within the U.S. Attorney General’s Office to challenge state laws in conflict with the “minimally burdensome national standard” that Trump has called for. While the new task force has not filed any lawsuits against states, the White House has stymied AI legislation or successfully pushed changes in Utah and Florida.
Republican state Sen. Joe Nicola of Grain Valley said he is in talks with the White House about a bill he’s sponsoring that would establish AI’s “nonsentience” and specify that liability for harm caused by AI must rest with a person or organization.
Nicola’s bill encountered opposition in the Senate from Republicans who cautioned it could run afoul of Trump and jeopardize $900 million in federal funding for rural high-speed internet development awarded to the state.
Nicola told The Independent that while he is changing the portion of his bill dealing with AI developers’ liability in response to White House feedback, a section establishing the sharing of deepfakes as a felony had not changed.
A bill sponsored by Republican state Rep. Scott Miller of St. Charles, which would also establish AI liability provisions, passed unanimously out of a House committee in March. The bill would absolve AI companies of criminal liability if they adhere to the federal AI Risk Management Framework established by the National Institute of Standards and Technology.
That, Miller said, could attract business to Missouri.
“It actually creates an appeal for people to relocate their AI business here, and the framework that we’re using is the same framework that every federal contractor would have to be compliant with anyway,” Miller said.
Lawmakers who sponsored parts of Hausman’s bill emphasized that Missouri has lagged behind other states in passing AI legislation.
Protections against deepfakes, said Republican state Rep. Bill Lucas of DeSoto, are “overdue.”
“We are living at a time when technology is evolving faster than our laws,” Lucas said. “Anybody with the basic tools can create and share harmful digital images that can destroy reputations and careers in seconds.”
Republican state Rep. Melissa Schmidt of Eldridge told The Independent that while lawmakers behind the bill will “yield to whatever we’re asked to yield to” by the federal government, they are determined to move forward.
“We’re trying to even just get something foundational passed into law that we can come back next year and build off of, because right now there’s nothing,” Schmidt said. “We are behind in this conversation.”
At least 26 states already have laws on AI-generated intimate depictions. And 46 states have some kind of AI legislation in effect, covering areas from the use of AI in political advertising to AI pricing algorithms.
Democratic state Rep. Elizabeth Fuchs of St. Louis voted for Hausman’s bill but asked the Senate to add a sunset clause.
“With AI technology advancing every day, we are legislating based on the best evidence that we have here today,” Fuchs said. “As AI continues to advance, I believe we owe it to the public to revisit this.”
During debate last week, Hausman said she expects lawmakers to amend the legislation in future years.
“I don’t think this bill is going to look the same in five years,” Hausman said. “I think it’s going to completely change, and I hope it does because we need to change with technology.”
We should not have people that are ignorant to technology legislating technology. It all boils down to security theatre and the erosion of rights.
ReplyDelete