Raleigh AI Coalition prioritizes training and data protection for Copilot launch

Raleigh City Council is taking significant steps to integrate artificial intelligence (AI) into its operations while prioritizing data security and employee training. During the March 18, 2025, work session, council members discussed the rollout of Microsoft Copilot, a tool designed to enhance productivity across various city departments.

A key decision made during the meeting was the proposal to mandate training for employees who will use Copilot. This training will focus on "prompt engineering," ensuring that staff understand how to effectively interact with AI to achieve desired outcomes. The council emphasized the importance of creating a clear AI policy and a guidebook that outlines best practices and potential pitfalls, particularly regarding the handling of proprietary information.

The council's discussions also highlighted the formation of a governance committee responsible for overseeing AI implementation. This committee, composed of members from the city manager's office and department heads, will prioritize how AI can best serve the community while adhering to safety protocols.

Another critical topic was the establishment of guardrails to protect sensitive data. Officials noted that any information fed into AI systems must be carefully managed to prevent the exposure of proprietary data. The council is considering mechanisms to alert users when they are about to input sensitive information into open models.

Additionally, the environmental impact of AI technology was acknowledged. The council is beginning to explore how the energy consumption of AI systems aligns with Raleigh's climate action goals. While discussions are in the early stages, there is a commitment to evaluate the sustainability practices of technology providers.

As Raleigh moves forward with AI integration, the council is focused on ensuring that all measures are in place to protect community data and enhance operational efficiency. The next steps will involve finalizing training programs and policies before a broader rollout of AI tools. This cautious approach aims to balance innovation with the responsibility of safeguarding public information.