Legacy Systems and Data Quality Challenges in AI Implementation: Five Recommendations for Local Government

Posted on September 8, 2025


Stack of computer punch cards fanned downward on a table. They are white, gold, and pink.

Today’s Morning Buzz is by Dr. Bill Brantley, President and Chief Learning Officer for BAS2A Inc., a state and local government consultancy. Connect with Bill on LinkedIn.

  • What I’m reading: “There’s Got to Be a Better Way: How to Deliver Results and Get Rid of the Stuff That Gets in the Way of Real Work” by Nelson P. Repenning and Donald C. Kieffer
  • What I’m watching: “Star Trek: Strange New Worlds”
  • What I’m listening to: “The Andy Beshear Podcast”

Artificial intelligence (AI) is now being used by cities and counties to improve services and manage budgets. However, many local governments struggle to integrate AI due to outdated systems and poor data quality. These challenges can be overcome with strategic planning. Even with limited resources, governments can get their data ready for AI. Here are five practical tips to address these issues.

  1. Start with a Data Inventory and Quality Assessment

Prior to implementing any AI solution, local governments should obtain a comprehensive understanding of their existing data environment. Frequently, agencies initiate pilot projects without fully recognizing the scope, storage methods, or usability of available information.

  • Conduct a thorough data inventory across all departments to determine which systems contain financial, human resources, permitting, or service-related records.
  • Evaluate data quality by identifying duplicates, incomplete fields, or inconsistent formatting. For instance, verify if addresses are standardized and dates are entered uniformly.
  • Identify redundancies and gaps. It is not uncommon for various systems to house slightly different versions of identical records.

This process establishes an essential baseline and reveals immediate opportunities for improvements, such as addressing evident errors that can quickly enhance data reliability and stakeholder confidence.

  1. Prioritize Interoperability with Low-Cost Integration Solutions

Fully replacing legacy systems is usually impractical due to budget limits. Prioritize interoperability so older and newer platforms can communicate. Use APIs for data sharing, middleware to integrate information from various sources, and cloud-based data lakes to centralize data securely. A gradual approach helps local governments modernize without major disruptions while preparing for AI solutions.

  1. Establish Clear Data Governance Policies

Technical solutions alone can’t guarantee AI readiness; poor governance still poses risks. Local governments should set clear rules for data collection, management, and sharing.

  • Form a data governance team to establish standards and resolve issues.
  • Implement metadata, naming, and version policies for consistent, reliable data.
  • Define data ownership to ensure accountability.

Effective governance prevents fragmentation and makes AI tools trustworthy.

  1. Invest in Staff Training and a Data-Driven Culture

Technology isn’t a complete solution; local government staff need skills for effective data use.

  • Train employees in data management basics: cleaning, labeling, and secure storage.
  • Build data literacy so staff can assess dataset quality before analysis.
  • Highlight wins where better data improved service, like faster processing or more accurate reports.

Staff value clean, interoperable data when they see its impact.

  1. Use Pilot Projects to Build Momentum and Demonstrate Value

Handle legacy systems by demonstrating AI’s value with incremental wins. Launch a pilot project focused on a targeted issue, like automating reconciliations or streamlining call center triage, to show how improved data boosts outcomes. Share what you learn to inform future modernization steps. Pilots help secure funding, as visible successes encourage leaders to support larger infrastructure upgrades.

Moving Forward on AI

Legacy systems and poor data quality hinder AI adoption in local government, but these issues can be addressed. Starting with a data inventory, prioritizing interoperability, establishing governance, training staff, and launching pilot projects help break data silos and enable innovation. The aim is sustainable AI integration, improving services and resource allocation. Leaders who tackle these barriers now will advance responsible AI use in their communities.

Close window