1. What Problem Are We Trying to Solve?

It’s easy to be impressed by new technology. But agencies should start by clearly defining the specific operational problem they want to address.

If you can't clearly articulate the problem, the solution — no matter how advanced — may not serve your needs.

‍

2. Is the System Transparent and Explainable?

AI systems used in public safety must be understandable. If an officer, supervisor, or citizen asks how a recommendation or decision was made, your agency needs an answer.

Ask vendors to explain — in plain language — how their system works, and what factors it relies on.

‍

3. What Happens When the System Fails?

No system is perfect. Agencies must understand what failure looks like — and who is accountable when errors happen.

Be sure to ask vendors for their incident response plans, fallback protocols, and how human oversight is maintained.

‍

4. How Does the Vendor Protect Data and Privacy?

Public trust depends on protecting sensitive information

Agencies must evaluate how an AI tool collects, stores, shares, and secures data — including any third-party access or ownership rights.

‍

5. Does This Fit Within Our Policy Framework?

Before buying any tool, agencies should assess whether the AI system aligns with their internal policies, civil rights obligations, and governance standards.

If your agency does not yet have an AI policy, creating one before procurement is critical.

‍

Final Thought

Adopting AI in public safety is not just about what a tool can do — it’s about what your agency can defend, explain, and trust.

Taking time to ask the right questions today can protect your officers, your community, and your agency’s future.