LOUISVILLE, Ky. — Artificial Intelligence could help states deliver services more efficiently, but it also threatens to displace workers and make it easier for bad actors to sow election misinformation.
Those were among the key takeaways Monday as state lawmakers from across the country convened here for the first official day of the National Conference of State Legislatures’s annual legislative summit.
AI is a featured topic at the summit this year. The at least eight sessions devoted to the issue reflect intense interest on the part of state policymakers as they seek to both promote the fast-evolving industry and establish guardrails.
Monday’s AI sessions were sometimes standing room only as lawmakers, staffers, industry representatives and others packed the rooms.
Colorado Senate Majority Leader Robert Rodriguez (D), who led the push to get a first-in-the-nation comprehensive AI regulation law passed, said state legislators have a chance to learn from past mistakes in trying to regulate consumer data privacy.
“I make an argument that we were too late on privacy in the United States and the longer we wait on AI, the more they’re going to say well this is how we’ve always done it,” Rodriguez told Pluribus News after moderating a panel on AI’s opportunities and challenges.
Rodriguez’s event featured Evi Fuelle, global privacy director at Credo AI, which helps companies safely adopt the technology. She urged lawmakers to regulate the fast-moving industry and said rules of the road will benefit, not harm, innovation.
“Businesses need certainty … to adopt this technology at scale,” Fuelle said. “[That] requires standards, it requires guardrails — things that businesses can point to [and] say, ‘I have met the criteria for what good looks like, and I’m being accountable and transparent.’”
Read more: AI takes center stage at NCSL summit
Momentum appears to be building for more states to follow Colorado’s lead and pass AI regulation next year. Three members of a bipartisan, multi-state AI working group told Pluribus News on Monday that they hope to see comprehensive legislation introduced in a dozen or more states in 2025.
“In the absence of federal legislation on algorithmic discrimination, how can we come together as states and adopt the same standards simultaneously, and so we’re currently working to put together that coalition,” said Maryland Sen. Katie Fry Hester (D), a member of the working group who co-chairs a joint committee in the Maryland General Assembly that has jurisdiction of AI policy.
Hester said success for her would be if California, Maryland, New York and Texas all passed laws with similar frameworks. Several AI bills are still pending in the California legislature, which returned Monday from summer recess.
State lawmakers introduced AI bills in at least 40 states this year, according to NCSL tracking. While comprehensive regulation proved a heavy lift, legislators in more than a dozen states, including New Hampshire, passed laws in 2024 regulating AI-generated election deepfakes.
New Hampshire Secretary of State David Scanlan (R), appearing on a panel focused on AI and elections, recounted a deep fake robocall in January that used President Biden’s voice to urge voters not to participate in the Granite State’s presidential primary. The Democratic political consultant behind the robocall now faces civil and criminal penalties
“It’s not new that campaigns use extreme methods sometimes to get their messaging across, you might call them dirty tricks … and this really was a form of that,” Scanlan told the audience.
New and cheap generative AI tools make it easier to produce fake voices, images and videos, amplifying the risk of election deepfakes.
Cait Conley, senior adviser to the director of the federal Cybersecurity and Infrastructure Security Agency, said states play an important role in setting the record straight in the face of disinformation campaigns, and investigating and prosecuting attempts to interfere with elections.
“Just like every other technology out there, there are going to be malicious actors who are going to look to leverage these capabilities for their own ends,” Conley said.
State officials are also concerned about AI’s effects on the workforce. Connecticut Sen. James Maroney (D), who is vice chair of the NCSL’s AI task force, cited a survey that found that 66% of business leaders said they would not hire someone lacking AI skills. He also pointed to another survey that found most employers currently use AI to screen resumes or plan to in the future.
“It’s important that we’re assessing these tools for bias … and making sure that we are providing training opportunities,” said Maroney, who advocated for state regulation and said Connecticut is offering a free online AI academy to help workers retrain.
The NCSL Summit kicked off with a general session featuring Jennifer Pahlka, a senior fellow at the Niskanen Center and former U.S. deputy chief technology officer.
She urged state leaders to use AI to reduce the time state employees spend navigating complex regulations with the goal of improving services to the public.
“If you use AI for that regulatory simplification … you’ll be burdening your folks less, so that they can do more of the right things that you care about on behalf of the people in your state,” Pahlka said.