ICMA authors propose a scoring matrix for vetting AI tools
As reported by ICMA's PM Magazine, a Leadership ICMA capstone project surveyed 28 local government professionals and found that 82% already use some form of AI at work. Uses include drafting council agendas and reports, translating and transcribing meetings and documents, and website navigation help. Respondents also named unclear policies and governance, data privacy and security worries, accuracy problems, and a lack of training.
The authors argue that traditional IT procurement is not always well suited to AI and that adoption is often happening through individual experimentation. Their answer is the AI Matrix, a web-based scoring tool that assesses functionality, workforce impact, human interaction, ethics and data security on a 30-point scale. A score above 16 counts as a pass. The sample is small, so treat the survey figures as indicative rather than representative.
For staff who prepare agendas, minutes and records, the practical issue is that tools touching public documents are often adopted without a consistent review. Questions about data handling, accuracy and human oversight matter more when outputs feed official records.
Practical takeaway: inventory which AI tools your office already uses, then ask whether any shared checklist, scoring approach or policy covers them. Even a simple common set of questions would reduce inconsistent practice across departments.