IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Vermont Bill Could Create a Permanent AI Commission

In a recently proposed bill, state lawmakers are asking to create a permanent commission to oversee and support the responsible use of artificial intelligence technology among state agencies.

Artificial Intelligence
In Vermont, a recently proposed bill looks to create a commission to oversee the ethical use of AI technology within state government.

The idea for the commission comes from a recommendation by the state’s Artificial Intelligence Task Force, which was operational from September 2018 through January 2020.

“On Jan. 15, 2020, the Legislature received a report from the task force recommending the state establish a permanent commission,” Rep. Brian Cina said. “The Committee on Appropriations heard testimonies and narrowed down the scope of the bill to create a permanent commission, along with a code of ethics and a public record of AI being used throughout the state.”

As for the bill itself, Cina explained, two bills were merged into one to create the current version of the legislation. The first bill aimed to create an index of current AI technology used within the state. The second bill looked to develop a code of ethics and support the responsible use of AI technology among state agencies.

“In Vermont, it is unclear at this point where we are using AI,” Cina said. “There is no inventory that exists.”

The idea of creating an inventory specifically came after law enforcement asked the state Legislature to issue an exemption for the use of facial recognition technology in 2020, Cina explained.

At the time, recently passed legislation banned law enforcement from using facial recognition technology due to racial justice concerns.

The exemption, according to Cina, was for law enforcement to use facial recognition technology to identify offenders and victims in child sex abuse cases.

A search warrant and other protections, including a human-reviewed decision-making system, would be in place to make sure there are as few discrepancies as possible in identifying individuals.

As for the new legislation, he said, one of the challenges may be getting enough financial support.

“There’s always an issue or concern when it comes to spending money,” Cina said. “The counterargument is that it’s an investment that saves the state money by ensuring that AI is being used wisely.”

“If we do end up using it in a harmful way, we could end up with a million-dollar lawsuit,” he added.

As for how state governments are currently using AI technology, most organizations have turned to using chatbots and robotic process automation, according to Amy Glasscock, program director of innovation and emerging issues at the National Association of State Chief Information Officers (NASCIO).

“The technology had gotten to a point where states felt comfortable using it, and there was a strong business case for using it,” Glasscock said.

Despite the increased comfort level with the technology, some challenges still persist around its implementation; privacy is one such concern.

“That is one of the big reasons states are hesitant to use it, especially when it comes to image recognition,” Glasscock explained. “States are really starting to put a big focus on privacy the same way that they were putting a focus on cybersecurity a decade ago.”

One way states are doing this is by hiring chief privacy officers. According to Glasscock, approximately 20 states have incorporated the new position into their organizations.

Another concern is incorporating AI and other related technology into legacy infrastructure.

“Legacy infrastructure is also another big challenge for states in using AI because there is so much of it,” she added. “A lot of states haven’t been able to move forward with a lot of AI applications.”

That’s not to say it’s not doable. For states looking to incorporate this technology into their daily operations, Glasscock recommended having a strategic vision, making sure all data is organized and standardized and to avoid looking for a business case to use new technology like AI. Instead, she suggested that identifying a business case first and then figuring out the best tool to use is a more practical approach.

“Every year, we ask state CIOs, what emerging technology do you think will be most impactful in the next three to five years,” she said. “AI and (robotic process automation) have been the top answer for a couple of years, so I think it’s going to continue to grow.”

Editor's note: Amy Glasscock's title was incorrect in the earlier version of this story.
Katya Diaz is a staff writer for Government Technology. She has a bachelor’s degree in journalism and a master’s degree in global strategic communications from Florida International University.