The council AI readiness checklist

Leadership and purpose

• We can name the two or three workloads where staff time is lost to admin.Re-entering notes, reformatting reports, chasing status. If you can't name them, start there.
• An executive sponsor is willing to own a small, measured pilot.A pilot with no sponsor tends to stall after the first demo.
• We know whether the goal is cash savings or recovered capacity.Most councils use recovered hours to clear backlog, which is a service gain, not a budget saving. Decide which before you measure.

Work and measurement

• At least one workflow is high-volume enough to measure before and after.A pilot needs a baseline. If you can't time it today, you can't prove an improvement later.
• The workflow already has a named owner and a record of the work.Job records and system counts make better baselines than staff estimates.

Risk and trust

• A human will review every AI-drafted output before it is submitted.The person remains the author of record. This is the single most important control.
• We know which steps must not be automated.Anything feeding prioritisation, safety or legal decisions needs extra scrutiny and testing first.
• We have checked what data staff are allowed to put into which tools.Settle this before the first session, not after.

People and tools

• Staff have access to tools the council already licenses.Check what you already have access to before buying anything.
• Local businesses and residents are part of the plan.A council's embrace of AI opens doors across its whole area. Consider an Economic Development session alongside any internal work.
• Field crews have a plan for patchy mobile coverage.Offline capture needs sorting out before any rollout beyond the depot.

Reading your result. Most boxes ticked: you are ready to talk about scoping a pilot. About half: close the gaps first, starting with a named sponsor and a measurable workflow. Few ticked: a short briefing is a better first step than a pilot.