Artificial intelligence (AI) is transforming the way governments serve their communities, from streamlining services to making smarter policy decisions. But as state and local agencies rush to adopt AI, a crucial truth is coming into focus: the quality of your data can make or break your AI initiatives.
The Data Dilemma in Government
A recent survey by the National Association of State Chief Information Officers (NASCIO) revealed that most state IT leaders feel their data governance is still in its infancy. Only a small fraction believe their systems are truly mature. Why does this matter? Because without strong data governance, even the most advanced AI tools can produce unreliable or biased results.
Doug Robinson, NASCIO’s Executive Director, points out that fragmented government structures have long made data quality a challenge. Now, as agencies look to integrate generative AI, the risks of poor data—like unfair resource allocation or wasted funding—are higher than ever.
Building a Data-Driven Culture
Improving data quality isn’t about finding a perfect dataset—it’s about investing time and resources to make your data better. Milda Aksamitauskas of the State Chief Data Officers Network emphasizes that the more agencies examine their data, the more issues they’ll uncover. But this process is essential for clarity and improvement.
One key step is appointing a dedicated data leader or committee. This ensures there’s always someone to turn to with questions about data quality or governance. Agencies with a Chief Data Officer (CDO) are often better equipped to lead data training and standardization efforts.
Ricki Koinig, CIO of Wisconsin’s Department of Natural Resources, highlights the importance of including diverse voices early in the data governance process. This helps prevent bias and ensures the data reflects the needs of the entire community.
Smart Procurement: Choosing the Right AI Partners
When it comes to adopting AI, agencies must be vigilant about who they partner with. Luis Videgaray of MIT Sloan School of Management warns that agencies need strong internal teams to vet vendors and scrutinize how their data will be used. Transparency is key—if a vendor can’t clearly explain how your data will be handled, it’s a red flag.
Data officers should be involved in procurement decisions, asking tough questions about data elements, formats, and potential biases. For smaller agencies, collaborating with neighboring governments can provide valuable insights and shared resources.
Setting Standards and Policies
Clear data policies and standards are the backbone of successful AI projects. Cities like Scottsdale, Arizona, have published detailed Data Service Standards to guide how data is collected, shared, and used. Indiana’s statewide data policy complements its AI policy, ensuring all agencies are aligned before launching new initiatives.
The Power of Data Audits and Readiness Assessments
Regular data audits help agencies understand where their data lives, who needs access, and what needs improvement. As more governments implement AI, these audits will become essential for identifying and fixing data issues before they derail projects.
Josh Martin, Indiana’s CDO, reminds us: “Garbage in, garbage out.” Without a clear understanding of your data’s quality and completeness, even the best AI tools will fall short.
Actionable Tips for Agencies
- Appoint a data leader or committee to oversee governance and quality.
- Include diverse perspectives in data discussions to minimize bias.
- Develop clear data policies and standards before launching AI projects.
- Involve data officers in procurement to ensure transparency and alignment.
- Conduct regular data audits to maintain data readiness.
A Practical Starting Point
For agencies just beginning their AI journey, experts recommend starting with AI-enabled chatbots or virtual agents. These tools rely on transactional data and can deliver quick, visible benefits while helping agencies build their data management muscles.
Key Takeaways
- High-quality data and strong governance are essential for effective AI in government.
- Appointing data leaders and including diverse voices helps improve data quality and reduce bias.
- Transparent procurement and clear policies protect data and public trust.
- Regular data audits and readiness assessments are critical for AI success.
- Starting with AI chatbots can provide early wins and valuable experience.
By focusing on these fundamentals, government agencies can unlock the true potential of AI—delivering smarter, fairer, and more efficient services for everyone.