The National Association of State Chief Information Officers (NSACIO) has released a new AI blueprint which features 12 key considerations to help states during the development of their own AI technology roadmaps.
“With the mass availability of generative AI (GenAI) tools and large language models in the last year, states are updating or creating new policies and road maps for artificial intelligence (AI),” NASCIO said. “As AI becomes increasingly integrated into the technology infrastructure of government agencies, an AI roadmap will emerge as an indispensable tool for states in the months and years ahead.”
NASCIO said that an AI roadmap not only facilitates the seamless adoption of AI but also enhances efficiency for already-strained state government workforces.
The organization highlighted a handful of benefits to creating an AI roadmap, including improving stakeholder and executive buy-in for AI initiatives; flexibility to adapt as the technology changes; efficient resource allocation and cost savings; risk management; and improved service delivery.
In the blueprint resource, NASCIO provided the 12 key considerations and info on how to implement them:
- Align AI initiatives to strategic drivers for the organization: Determine how AI fits into the overall goals of the state IT strategic plan. Identify the business case and overall strategic goals of the organization before deploying an AI tool.
- Establish governance and oversight processes: Absence of proper AI governance exposes states to potential risks such as data leakage, violations of privacy laws and erosion of citizen trust. To navigate these challenges, it is crucial to adopt established AI governance frameworks.
- Inventory and document existing AI applications: Discover the extent of AI tools employed by agencies, both knowingly and unknowingly.
- Address data quality and sourcing: Prioritize data governance and classification to ensure the highest quality data available is used. Evaluate data sources, mindful of potential biases.
- Collaborate with stakeholders and industry partners: Create an advisory board or task force with key internal stakeholders. Build industry partnerships to leverage expertise and innovation.
- Assess privacy and cybersecurity risks of AI adoption: Consider performing privacy and security impact assessments for new AI technologies.
- Infrastructure and technology: Assess the current state of the technology infrastructure and identify areas for improvement. Legacy infrastructure has been a common roadblock for states.
- Create acquisition and development guidelines: Develop best practices and guidelines for acquisition/procurement, development, and operation of secure AI systems. Update procurement language as needed to cover AI and generative AI concerns.
- Identify potential use cases: Research potential use cases of AI by learning from other government entities, organizations, and AI leaders.
- Expand AI workforce expertise and training: Identify and amplify existing staff expertise, recruit interns and staff, partner with local educational institutions and provide training and educational opportunities for employees.
- Create guidelines for responsible use, ethics, and transparency: Ensure that users of AI systems are informed about the risks associated with discrimination and bias.
- Measure and communicate effectively: Have clear metrics in place to measure progress and success of AI initiatives.