Alabama AI Task Force reviews preliminary recommendations ahead of November report

A person using an artificial intelligence chatbot. An Alabama task force in Montgomery on Thursday reviewed preliminary AI policy recommendations, including frameworks, risk management and AI literacy, ahead of a report to be released in November. (Getty Images)

On Thursday, a state task force created to regulate the use of artificial intelligence (AI) in state government reviewed preliminary policy recommendations ahead of a report to be released in November.

Task force members in four separate working groups outlined possible ways to better use AI in government, including definitions of AI policy frameworks; improved risk management practices and improved AI literacy across government agencies.

“I think there are a lot of good ideas and frameworks. Obviously, you can’t solve everything through this working group, but there are good, good straw man templates that we can provide to agencies,” said task force chairman Daniel Urquhart, secretary of the Alabama Office of Information Technology.

SUBSCRIBE: GET MORNING HEADLINES DELIVERED TO YOUR INBOX

Members of the task force frequently discussed Generative AI, or GenAI, a type of artificial intelligence that can create new content such as text, images or sound, according to McKinsey, a management consulting firm.

The policies and governance working group has proposed developing a comprehensive set of definitions to standardize AI terminology within government. Presenting on behalf of the group, Mike Owens emphasized the importance of a risk management process designed for AI and proposed an oversight board that would ensure AI initiatives comply with state regulations and encourage compliance.

“Whether it’s security systems management or application development, have those skills validated by the assurance board,” Owens said.

Laura Crowler, from the policies and governance working group, said they would propose policy recommendations, not legal changes.

The workforce, education and training group discussed efforts to build AI knowledge among government employees. Roger Bowman proposed the development of an introductory AI course developed in partnership with academic institutions and professors from the University of Alabama and Auburn University working in the area.

“We’ve approached this in two ways: First of all, we’ve looked at some of the other introductory AI training courses that other states have put together… The other area where we’re approaching training the workforce is we’re looking at an option to highlight some of the GenAI functionality and make it available for consumption and for users to test it out,” Bowman said. and publish it where they can see how the capabilities work,” he said.

A working group evaluating responsible data governance has proposed guidelines on data ownership and classification of data for use with AI systems that minimize risks of error or bias. The group also recommended that institutions develop a reference guide to ensure that sensitive or restricted information is handled ethically.

“We want to help make sure that when GenAI systems are used, citizen data (if used with GenAI data) is protected,” said Aaron Wright, who presented for the working group.

Presenting on behalf of the responsible and ethical use of AI group, Willie Fields discussed the need for privacy, bias monitoring, and accountability in the use of AI. Fields also said they will not propose legislation, but instead focus on policy recommendations. Fields also noted accountability, especially in determining who is responsible for AI-focused decisions.

“Who is responsible for the data coming from these GenAI systems? Ultimately, that person is the head of the agency, but we need to help that person be ready to sign off or authorize the system to go into production,” Fields said.

The task force is expected to finalize its recommendations at its next meeting in late October and the report will be presented to Gov. Kay Ivey in November.

DONATE: SUPPORT THE NEWS YOU TRUST