Arizona Pioneers Practical Uses for Generative AI

PHOENIX (Nov. 20, 2024) - Arizona state government is testing Generative AI to improve public safety and government efficiency to lay the foundation for further use in the public sector.

In support of testing and adopting, the State of Arizona is dedicated to documenting best practices for the safe, effective and responsible use of Gen AI throughout state government. Governor Hobbs recently announced the creation of an AI Steering Committeeto help inform and drive the future deployment of artificial intelligence in the state. The newly formed steering group will explore ways Gen AI can benefit the public and shape policies that promote transparency and ethical and responsible use of the technology.

Since the initial publication of the statewide Gen AI policy and procedure in March 2024, the Arizona Department of Administration issued updates to address rapid advancements in the technology. These updates to the state's Generative AI Policy (P2000)and Use of Generative AI Procedure (2000PR) are available to the public.

The guiding principles of the state's Gen AI policy include:

• Empowerment - Supporting our workforce to deliver services efficiently, safely and equitably to the public.
• Transparency and Accountability - Building trust and enabling collective learning.
• Fairness - Seeking to ensure that our use of technology uplifts communities equitably.
• Privacy and Security - Responsible experimentation that maintains control and respects privacy and security.

Major changes to the statewide Gen AI policy and procedure include:

• Updates on the role of the State Data and Analytics Office, which was established in May 2024 to advance data efforts and support Gen AI usage across the state.
• Emphasis on the importance of data governance policies and data readiness.
• Expanded details on agency and employee responsibilities regarding responsible use, data protection, transparency, accountability, security and privacy.

"These testing applications for Gen AI and associated updates to the statewide policy and procedure are a reflection of how fast this area of technology is developing and advancing," said State of Arizona Chief Information Officer J.R. Sloan. "Our policy seeks to provide the guidance and guardrails that enable the safe, responsible and effective use of technology that supports the productivity of our employees in serving the people of Arizona. We will continue to keep pace with developments to ensure that we benefit from these technologies to further serve every Arizonan."

The State of Arizona is already applying several Gen AI tools to improve government for Arizonans and public servants alike:

• The Arizona Health Care Cost Containment System has leveraged Gen AI and responsive technology to make health care simpler for patients and providers. In the summer of 2023, AHCCCS introduced its new chatbot named SAM to help members get fast answers to more than 40 common questions. In January 2024, AHCCCS introduced Gen AI to assist with cyber security and fraud prevention. Gen AI is also at the center of an upcoming fraud prevention effort that will automatically flag and learn concerning billing practices that could indicate fraudulent activity and patterns.
• In January 2024, Arizona State University became the first higher education institution in the U.S. to collaborate with OpenAI. The ultimate goal of this collaboration is to discover paths toward optimizing ChatGPT Enterprise for faculty and staff to enhance teaching and learning. Starting in February 2024, ASU invited faculty and staff to submit suggestions on the practical use of Gen AI. The key areas of concentration include enhancing student success, forging new avenues for innovative research and streamlining organizational processes. At the forefront of higher education, ASU continues to lead and innovate by responsibly leveraging technology solutions. Through the collaboration, ChatGPT Enterprise ensures that the platform places user privacy as a top priority by implementing enterprise security measures that protect user data from digital threats.
• Starting in April 2024, the Arizona Strategic Enterprise Technology Office worked with three vendors (Google, AWS and Microsoft) to develop sandbox environments for agencies interested in learning and testing the capabilities of Gen AI. Agencies had the opportunity to select which of the three vendor sandbox environments they wished to use and to conduct tests on agency projects not yet moved into production environments. Training and guidelines from each vendor were made available to agencies. The ongoing sandbox testing includes the participation of the Arizona Department of Revenue, the Arizona Department of Environmental Quality, the Arizona State Retirement System, the Arizona State Board for Charter Schools and the Arizona State Board for Private Postsecondary Education.
• In September 2024, the Arizona Department of Administration conducted a four-week pilot program on Gemini for Google Workspace with the participation of 203 users and nine agencies. The state's goal was to explore ways to automate routine tasks and improve efficiencies to allow public servants to better serve Arizonans. The results of the pilot program suggest that by utilizing Gemini, employees could potentially increase their productivity by 2.5 hours per week, helping employees summarize information, automate charts and graphics, take notes, format, write code and much more.
• In October 2024, the Arizona Department of Public Safety launched a partnership with TRULEOand Arizona State University to use artificial intelligence to conduct a field study on reviewing body camera footage. TRULEO converts body camera video to text and automatically redacts police officers' and community members' personally identifiable data. The tool detects and logs things like introductions, explanations, and de-escalation attempts. Use of this AI expands law enforcement's capacity to review footage, improves retention and morale, and reduces bias.