Calvert County adopts AI policy and forms advisory committee to vet tools, set guardrails

The Board of County Commissioners voted to approve an artificial intelligence use policy and establish an AI advisory committee that will advise the county administrator on AI initiatives, authorized use cases and training.

Technology Services staff presented the policy’s key elements: a risk matrix to assess data sensitivity, mandatory cybersecurity vetting before an AI tool can be used, departmental discretion for whether a tool is allowed in a given division, mandatory training for staff who receive licenses, and a quarterly advisory committee made up of representatives from the county administrator’s office, legal, HR, communications, and Technology Services. Change-management and communications manager Jonathan Lee explained the rollout and said the policy emphasizes “human in the loop” controls and staff responsibility for AI outputs.

Commissioners asked about existing use and oversight: staff said the county already uses commercial tools, noted roughly 101 ChatGPT licenses in use within departments and five to six paid professional licenses held by Technology Services, and discussed the option of buying managed licenses to retain logs and create department-specific GPTs. County Administrator Mark Wallace described early productivity benefits and cautioned that content must be verified before release.

The motion to approve the AI policy and form the advisory committee was moved, seconded, and approved by voice vote. The policy requires cybersecurity vetting of tools before committee review and gives department heads final say about use within their departments.

Next steps: Technology Services will finalize committee membership, publish policy guidance and training requirements, and begin licensing discussions to improve oversight and logability.