IE 11 Not Supported

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

Arizona’s AI Policy Is Evolving Along With the Technology

The state first published its policy and procedure for generative artificial intelligence in March. Since then, officials made several updates to address the changing needs this type of technology creates.

Digital illustration of a hand moving markers on a map from point A to point B
Shutterstock
The state of Arizona published its policy on generative artificial intelligence (GenAI) in March, but as the technology and the state’s applications of it have evolved, the state's policy has changed, too.

State and local governments have implemented a variety of policies related to AI to fill a regulatory gap at the federal level. In September, Arizona CIO J.R. Sloan told Government Technology that AI is in the “productivity phase,” explaining that the state is focused on establishing both policy and procedure for GenAI.

This week, the state announced several updates to the statewide GenAI policy and procedure (P2000 and 2000PR, respectively), which were first published in March. The updates were officially enacted Oct. 24, according to an email from Bryan Beach, communications and digital content manager within the Arizona Department of Administration (ADOA)’s Strategic Enterprise Technology (ASET) Office.

There are three major changes to the GenAI policy. The first entails updates on the State Data and Analytics Office’s role; this office was established in May to advance data work in support of GenAI. The second is added emphasis on the importance of data governance policies. The third is the addition of details on both agency and employee responsibilities related to GenAI applications, including responsible use, data protection, transparency, accountability, security and privacy.

The updates are intended to address rapid advancements in AI technology. According to Beach, these changes were made through a process in which ADOA consulted with agency CIOs. This involved sharing draft versions of the policy and procedure with proposed updates, and incorporating input from agency CIOs.

Last week, Arizona Gov. Katie Hobbs announced the creation of an AI Steering Committee to help inform future AI deployment and identify potential applications for its use. This committee, Beach noted, will advise on future GenAI policy and procedure updates.

The policy changes support the state’s use of GenAI in state government; several such state applications were highlighted in the announcement.

For example, in September, the ADOA conducted a four-week pilot involving 203 users and nine agencies. This initiative involved the use of Gemini for Google Workspace, resulting in a demonstrated productivity increase of 2.5 hours per week.

Another example started in April, in which ASET worked with three vendors to develop GenAI sandbox environments for agencies to experiment and learn safely. During this process, each vendor made training and guidance available to agencies. This testing is ongoing.

The state has even used GenAI in health care; the Arizona Health Care Cost Containment System has used a GenAI-powered chatbot since 2023 to help answer customer questions. In January, the agency also started using GenAI to assist with cybersecurity and fraud prevention. Increased use of GenAI for fraud prevention related to billing practices is forthcoming.
Julia Edinger is a staff writer for Government Technology. She has a bachelor's degree in English from the University of Toledo and has since worked in publishing and media. She's currently located in Southern California.