BOX 1
STRATEGIES FOR INTEGRATING AI TECHNOLOGIES

1. Foundations and Governance (Why use AI and how to ensure its responsible use?)
   • Be Purpose- and People-Oriented: Begin by grounding AI initiatives in public values and a clear understanding of both human and organizational contexts. Use AI not only to solve well-defined problems but also to explore new possibilities, anticipate future needs and augment the capacity of public institutions to respond to complexity.
   • Engage the Public: Create opportunities for engaging the public through consultations, forums, surveys, feedback mechanisms, and participatory design processes to promote transparency, increase public trust, and enhance decision-making.
   • Build Proportional and Iterative AI Governance: Develop internal governance policies that scale with the scope and risk of the specific AI use. For high-impact or public-facing systems, adopt frameworks such as the National Institute of Standards and Technology’s (NIST’s) AI Risk Management Framework (RMF) (2023) or the National League of Cities’ Artificial Intelligence Demystified—AI Toolkit for Municipalities (2024). For low-risk tools and technologies, prioritize streamlined, values-based guidelines that promote responsible use without stifling innovation.
   • Participate in and Help Shape Emerging Collaborative Frameworks: Engage with existing and developing federal–state–local collaborations on AI governance that can provide technical assistance, shared toolkits, legal frameworks, and coordination infrastructure to help all levels of government adopt equitable and interoperable AI practices, regardless of local capacity or political alignment.
   • Develop Tiered AI Procurement Guidance: Develop tiered AI procurement guidance that balances risk, use case, and vendor capacity. Encourage ethical and legal vetting for high-impact AI tools and technologies while offering streamlined pathways for use and innovation for products from small businesses and civic technology groups. Provide centralized support or shared services to help governments apply this guidance.
2. Planning and Scoping (How to get started responsibly?)
   • Conduct Feasibility Assessments and Workflow Mapping: Where possible, conduct initial feasibility and workflow assessments that center the user experience to help determine what needs to be built or procured, why, and how. Prioritize scoping not just technical capacity but also oversight and governance needs. Provide shared frameworks or templates to reduce the burden on users.
   • Scope Internal Capacity: Identify staffing, technical infrastructure, and oversight roles necessary to support long-term operation, maintenance, and responsiveness.

3. Design, Development (if internal), or Selection (if procuring externally) (How to align design and development with purpose and use?)
   • Align the Problem Definition with Context, Goals, and Technical Design: Before adopting AI, refine the problem statement, incorporating input from all stakeholders. Explicitly enumerate the list of design choices that need to be made to develop or procure the system and match those choices to the specific deployment settings.
   • Define Evaluation Criteria and Assess System-Level Impacts: Define evaluation criteria that reflect the social and organizational context, and before deployment, set baselines with which to track progress over time. Evaluate the whole system—not just the AI model—including downstream users and the system’s impact on people.
   • Establish Feedback Mechanisms: Ensure that feedback mechanisms, including measures for public perception and staff feedback loops, are established to identify potential challenges and opportunities.
4. Capability and Culture (How to build readiness?)
   • Build Internal Capacity and Competency: Enable effective adoption and implementation of AI systems by fostering a culture of adaptation, increasing internal capacity through training and hiring, targeting technical capacity, supporting collective and continuous improvement, and prioritizing leadership buy-in.
   • Use Partnerships with Stakeholders: Engage in collaborations with civil society, industry, or academic partners to gain access to cutting-edge technology, pilot programs, security strategies, and technical support.
5. Ongoing Accountability and Engagement (How to manage implementation and maintain trust?)
   • Establish Tiered Continuous Monitoring and Improvement Mechanisms: Develop tools to monitor and audit AI systems throughout their life cycle using a tiered approach that considers not only organizational size and resources, but also the level of risk posed by the AI use case. Develop evaluation plans that provide comprehensive measures, covering the entire life cycle from inception to pre-deployment testing, through real-world performance monitoring, and including pathways for system improvement or decommissioning.
   • Create and Sustain Advisory and Oversight Bodies: Establish advisory and oversight bodies with enforceable authority, clear mandates, effective broad representation, and established mechanisms for translating advice into actionable insights to increase transparency and accountability.