In San Jose, Calif., the city’s Information Technology Department issued guidelines for government use of generative AI, recognizing “the opportunity for a controlled and responsible approach that acknowledges the benefits to efficiency while minimizing the risks around AI bias, privacy and cybersecurity.”
Government Working Groups Examine AI Best Practices
San Jose details seven guidelines for generative AI use by local agencies, including making generative AI systems subject to public records access, double-checking AI results against multiple sources and participating in AI best practice workgroups.
Nationally, key organizations have established AI working groups. Through its Center for Best Practices, the National Governors Association assembled a list of state resources on AI to offer guidance to state governments. The National Association of Counties established an AI Exploratory Committee. Among the committee’s mandates is a mission to examine “consequences of AI through the lens of county government governance, operations, constituent services, innovation, public trust, privacy and security, and workforce productivity.”
In a recent blog post, Microsoft notes that Copilot is “set to redefine how government agencies work.” But the company also emphasizes its commitment to “responsible AI innovation” while securely broadening AI access to agencies that handle sensitive information.
Microsoft promises to introduce Copilot to its Microsoft 365 Government Community Cloud users in summer 2024. Working together, the company and state and local governments can prepare for this first major test of mass AI deployment by states, counties and cities.
