As a growing number of public agencies look to implement tools and programs that rely on artificial intelligence (AI), elected officials and senior agency leadership have important roles to play in promoting the responsible adoption and use of these systems. Effective governance and oversight of AI is critical for state and local governments to realize good value from their AI projects, increase efficiency, build and maintain public trust, avoid failed AI projects and public relations disasters, and improve constituent services. Most importantly, these measures help governments ensure that AI tools are positioned to assist agencies in delivering on their missions of improving constituents’ lives and supporting the public good.
This brief provides elected officials and senior leaders working in state and local government with a checklist of core recommendations specifically aimed at building government-wide structures, strategies, and processes to advance trustworthy and responsible use of AI in public benefits and services across five core areas:
Public Transparency and Stakeholder Engagement: Improve public awareness and understanding of AI by establishing public AI inventories, prioritizing public education about government use of AI, creating advisory councils with members of the public to inform agency AI decision-making, implementing mechanisms for meaningful feedback from the public, and instituting plain-language notices and explanations for affected individuals.
Accuracy and Reliability: Ensure that AI projects advance agency goals and combat AI-driven challenges by adopting acceptable AI use policies or guidelines, grounding the acquisition and use of AI tools in evidence-based decision-making, establishing minimum government-wide AI performance and testing standards and procurement criteria, implementing regular independent audits of AI tools (including post-deployment), building in requirements for human oversight and training, and prioritizing investment in AI talent.
Governance and Coordination: Promote cross-agency governance practices by adopting a government-wide AI plan and governance strategy, appointing a chief AI officer or equivalent senior leader, creating AI governance boards, establishing centralized emergency response protocols and AI incident reporting, engaging cross-functional staff in AI decision-making, establishing forums for government employees to provide input on AI projects, and incorporating responsible AI guidance into existing employee training and onboarding materials.
Privacy and Security: Identify and mitigate AI-related privacy and security harms by updating cybersecurity and data policies; establishing privacy and security protections in AI procurement; integrating chief privacy, information security, and data officers throughout AI decision-making; and prioritizing privacy and cybersecurity in employee AI training.
Safety, Rights, and Legal Compliance: Address the risks that AI systems may pose to the public’s safety and rights by integrating civil rights, risk, and legal officers throughout AI decision-making; establishing heightened risk management requirements for high-impact uses; and prioritizing legal compliance and identification and mitigation of AI harms in employee AI training.
Depending on the role and function of an elected official or senior leader, these recommendations may be carried out through actions such as policymaking, guidance, rulemaking, executive orders, legislation, city or county ordinances, or oversight. Critically, elected officials and senior leaders will need to build capacity within agencies to deliver on these five core areas.
State and local governments use various types of AI and automated systems — such as generative AI, agentic AI, predictive analytics, automated decision-making systems, and facial recognition — that each come with their own risk considerations. The recommendations included in this brief should be adapted to the type of AI in a given use case, with higher-risk applications needing heightened guardrails.
Coalition Urges Senate Not to Let Companies Waive Financial Regulations for AI
CDT joined AI Now Institute, American Civil Liberties Union, and several organizations dedicated to tech policy, consumer protection, and civil rights in a letter to Senate leadership and the Senate Banking, Housing, and Urban Affairs Committee opposing the “AI Innovation Labs” language in Sec. 10509 of the CLARITY Act.
As concern about risks and harms related to AI systems continue to grow, a growing chorus of policymakers, industry leaders, and advocates have called for independent AI assessments. This explainer provides an overview of recent proposals for third-party assessment in the United States, including state and federal legislation, executive actions, and industry proposals.
Having third parties assess AI systems might seem like common sense, but crafting effective policies toward this goal can be devilishly tricky. A poorly-constructed ecosystem for third-party assessment could easily fail to consider the most consequential mechanisms of risk, neglect the AI harms that most impact people, or do more to protect AI companies than people.
Not All Guardrails Are Created Equal: Comparing Content Safety and Copyright Filtering
As courts and policymakers work through questions about chatbot liability, they should be wary of analogies that flatten meaningful technical differences. Copyright filtering and safety intervention share real challenges around ambiguity and evasion, but they diverge in what each control must assess, how each manifests over the course of a conversation, and how much can be verified from the outside.
