Government Leaders Outline Responsible AI Priorities for State and Local Agencies

Government and industry leaders used the National Academy of Public Administration’s November 10, 2026, conference in Washington, D.C., to discuss how artificial intelligence could improve public services while preserving human judgment and public trust. The conversation focused on applications including permitting, legacy-system modernization, cybersecurity, and routine administrative work. For GovCon companies, these are areas to monitor as state and local agencies consider AI adoption, but the conference discussion itself was not a procurement announcement: it identified no solicitation, contract value, award, or procurement contact.

Tempe, Arizona, Mayor Corey Woods described the city’s experience adopting AI governance as employees began using dozens of tools. Tempe adopted an ethical AI policy in 2023, developed largely by city staff, including CIO Jared Morris and his team. Woods said the policy was intended to provide transparency about how AI is used and what safeguards are in place, explaining: “We needed to adopt an ethical AI policy, so they know we’re not just simply, you know, willy-nilly using any program.” For vendors, the example highlights that agencies may seek not only AI products, but also support for governance, transparency, and responsible implementation.

Permitting emerged as a practical service-delivery use case, particularly for small businesses that may not have resources to hire specialists to navigate government processes. Woods said, “Whatever we can actually do to help those folks get through the process a lot quicker, we want to do that.” Anthropic’s Thiyagu Ramasamy also highlighted the modernization opportunity: “There’s so much legacy code in the system that could be modernized. The permitting system could be really old. It’s much easier to actually rewrite it to modern technologies much, much faster.” This suggests potential interest in solutions that combine process improvement with modernization of aging applications, though agencies would still need to define requirements, conduct evaluations, and establish an acquisition path.

Panelists also stressed that AI adoption requires safeguards to evolve alongside the technology. OpenAI partnership manager Alexis Bonnell said organizations should revisit governance policies regularly, warning that if they are not updated “on a monthly basis,” agencies risk missing both new risks and new opportunities to serve the public. She emphasized keeping people in control and noted that carefully selected source materials and custom instructions can help tailor tools, while not replacing organizational safeguards. The procurement takeaway is that buyers should assess workforce readiness, human oversight, governance, and public trust alongside technical capability. The remarks signal a market area for vendors serving state and local government—not a currently identified acquisition.

• Tempe adopted an ethical AI policy in 2023 after city employees had begun using dozens of AI tools.
• Mayor Corey Woods identified permitting as a way AI could help residents and small businesses navigate government processes more quickly.
• Anthropic public-sector leader Thiyagu Ramasamy pointed to legacy-code modernization and permitting systems as potential AI applications.
• The conference discussion also identified cybersecurity, routine administrative work, service navigation, and research acceleration as areas of potential use.
• OpenAI partnership manager Alexis Bonnell urged organizations to revisit AI governance regularly as capabilities and risks change.
• Panelists emphasized human control, safeguards, workforce preparation, and public trust; AI tools do not eliminate the need for organizational governance.
• Vendors can align offerings with permitting and legacy modernization while addressing implementation, governance, and responsible-use needs.
• This is a market signal, not an active procurement notice: no solicitation, award, contract value, or procurement contact was identified.