States Should Move AI Pilot Programs from Siloed Tests to Statewide Deployment – Center for Data Innovation
Publish Date: 2026-06-04 15:37:00
Source Domain: datainnovation.org
-
AI’s Impact on Public Sector: AI is transforming industries, and state lawmakers are creating opportunities for its integration into government services to keep pace.
-
Emerging State Models: Despite numerous pilot programs, a gap exists between innovative, whole-government experimentation and isolated agency tests. Clear evaluation metrics are needed to avoid cycling through ineffective experimentation.
-
Utah’s Regulatory Sandbox: Utah led with the 2024 Artificial Intelligence Policy Act, establishing a regulatory sandbox to test AI tools under temporary relief, focusing on real-world performance data for future regulation.
-
Connecticut and Collaborative Oversight: Connecticut introduced an AI Responsible Use framework in 2024, setting standards for safe deployment and creating an AI Engagement and Enablement Lab with expert oversight.
-
Ohio’s Comprehensive Governance: Ohio’s 2024 AI Governance Framework created a multi-agency AI Council with a technology sandbox for safe experimentation, emphasizing ethical standards and privacy controls.
-
Texas’s Structured Pilot Programs: In 2025, Texas enacted an AI Governance Act establishing a regulatory sandbox with a dedicated AI Council and a testing window for diverse AI systems.
-
North Carolina’s Rapid Testing and Leadership: North Carolina’s Executive Order created an AI Accelerator with rapid-testing cycles and AI Oversight Teams to identify use cases, fostering streamlined public financial record management.
-
Best Practices for Policymakers: Policymakers should adopt sandbox models from these states, prioritizing whole-government experimentation, robust oversight, and centralized testing to responsibly integrate AI into public services.
