ATLANTA – A Georgia Senate study committee on Wednesday set a broad framework for determining how the state should regulate emerging artificial intelligence technology to protect the public without stifling innovation.

“(AI) will literally cure cancer,” Sen. John Albers, R-Roswell, the study committee’s chairman, said during the panel’s first meeting. “However, it also has the propensity to do great harm. … It’s going to impact and change things like never before.”

Several legislative committees held hearings on AI last year, and a bill was introduced in the Georgia House of Representatives during this year’s legislative session to criminalize the use of “deepfakes” generated by artificial intelligence to impersonate candidates in political ads. House Bill 986 overwhelmingly passed the House but died in the Senate.

On Wednesday, the new Senate study committee agreed on a broad range of policy areas AI will affect that need to be addressed in any legislation Georgia lawmakers come up with, including health care, public safety, education, and transportation.

Overlapping all of those categories is how to regulate AI in a way that ensures the technology is being used ethically and transparently. A House committee planning to begin meeting soon will also take up that issue, said Rep. Brad Thomas, R-Holly Springs, who was the chief sponsor of the deep-fakes bill.

Georgia could be among the first states to adopt regulations for AI. While the European Union’s Parliament adopted AI legislation last March, Colorado is the only U.S. state to have done so, Hayley Williams, director of the state Senate Office of Policy and Legislative Analysis, told the Senate panel.

Congress thus far hasn’t passed any AI regulations, she said.

“It’s a very complex universe to deal with and very difficult to regulate,” she said. “The reality is, the impact is too huge not to regulate.”

More: A ‘perfect tool’ to increase division: Augusta University professor talks TikTok ban

Williams said the EU’s Artificial Intelligence Act, which will take effect in 2026, regulates AI systems based on the risk they pose to the public. AI systems that pose an “unacceptable” risk are prohibited altogether, while systems considered to pose “minimal” risk are not regulated at all.

European companies that fail to comply face stiff fines, Williams said. Colorado’s law does not impose fines for non-compliance, she said.

Sen. Max Burns, R-Sylvania, said the study committee’s goal should be to foster innovation in the development of AI in Georgia with less emphasis on imposing restrictions like the EU model.

But Sen. Jason Esteves, D-Atlanta, said regulating AI systems to protect the public also must be an important goal.

“The primary function of government is to protect its citizens,” he said. “We should be ensuring we protect citizens from the potential impacts of AI.”

Albers said he plans to schedule seven or eight meetings of the study committee this summer and fall before the panel makes recommendations to the full Senate. The next meeting is set for July 17.