Arizona lawmaker uses ChatGPT to help craft legislation to combat deepfakes

PHOENIX — A Republican member of the Arizona House used ChatGPT to help craft legislation on artificial intelligence-driven impersonations, which was signed into law by Democratic Gov. Katie Hobbs earlier this week.

State Rep. Alex Kolodin used the AI software to help define “digital impersonation” in Arizona’s new law, which aims to regulate deepfake technology. House Bill 2394, which Hobbs signed into law Tuesday, gives Arizona politicians and other residents the ability to get a court order declaring that the person in the deepfake is not them.

It comes in the wake of a broader national discussion over digital impersonations. On Monday, actor Scarlett Johansson said that artificial intelligence company OpenAI used an “eerily similar” voice to hers for its new chatbot despite her having declined the company’s request for her to provide her voice. OpenAI announced it would stop using that voice, claiming "it was never intended to resemble" Johansson's.

Meanwhile, more states are introducing and passing legislation to deal with deepfakes ahead of the 2024 election, amid broad concern over the effect of AI-driven disinformation.

“I used it to write the part of the bill that had to do with defining what a deepfake was,” Kolodin said on his use of the software in the legislative process.

“I was really struggling with the technical aspects of how to define what a deepfake was. So I thought to myself, ‘Well, why not ask the subject matter expert, ChatGPT?’” Kolodin said.

The legislator from Maricopa County said he “uploaded the draft of the bill that I was working on and said, you know, please, please put a subparagraph in with that definition, and it spit out a subparagraph of that definition.”

“There’s also a robust process in the Legislature,” Kolodin continued. “If ChatGPT had effed up some of the language or did something that would have been harmful, I would have spotted it, one of the 10 stakeholder groups that worked on or looked at this bill, the ACLU would have spotted, the broadcasters association would have spotted it, it would have got brought out in committee testimony.”

But Kolodin said that portion of the bill fared better than other parts that were written by humans. “In fact, the portion of the bill that ChatGPT wrote was probably one of the least amended portions,” he said.

He argues that any shortcomings associated with using ChatGPT to write part of a law would also be present if humans take the reins. Kolodin said he didn’t see any pitfalls “that I don’t also see with relying on legislative attorneys to draft up legislation.”

Kolodin noted that he used ChatGPT sparingly in the process, employing it for a technical definition of “digital impersonation” while leaning on his experience as a lawyer and politician as well as the legislative process.

A representative for the governor’s office confirmed that Hobbs wasn’t aware Kolodin had used ChatGPT to help draft the bill she signed into law. That was an omission Kolodin admits was by design.

“I kind of wanted it to be a surprise once the bill got signed,” he said.

This article was originally published on NBCNews.com