Federal agencies are rushing into AI without cleaning house first
Federal agencies are rushing into AI without cleaning house first
Publish Date: 2026-05-18 17:20:00
Source Domain: www.nextgov.com
In Focus: The Overlooked Role of Knowledge Management in AI Deployment for Federal Agencies
Dr. Tori Reddy Dodla sheds light on the pitfalls federal agencies face when deploying AI-driven tools like Microsoft 365 Copilot without adequately addressing their existing knowledge management and governance frameworks. Dodla emphasizes that while the rush to deploy AI-driven productivity tools is palpable, the underlying security and data management issues are significantly more problematic. Over years of neglecting knowledge management practices, agencies are now grappling with inconsistent governance standards, resulting in critical documents and sensitive information being exposed to unauthorized users. Dodla’s experience highlights the risks: when Copilot surfaces information users already have access to, but often shouldn’t view, bad decisions may stem from outdated or incorrect data. The essence of Dodla’s argument lies in three imperative areas for federal agencies: re-establishing knowledge management as essential infrastructure, customizing data governance strategies specific to federal needs, and enforcing robust permission management at the site level. Effective preparation means avoiding costly, time-consuming cleanup efforts later due to the exposure Copilot may make visible.
Key Points:
- Knowledge management is often overlooked, leading to security risks when deploying AI tools.
- Customized data governance strategies are crucial for federal AI readiness.
- Effective permission management and site ownership are essential to ensure secure AI deployment.
- Agencies must enforce robust SOPs to prevent issues stemming from outdated or mismanaged data.
- Inadequate knowledge management can lead to leaks of sensitive information and erode public trust.
