China AI for City Government: A Study-Mission Framework

For a city-government delegation, the unit of comparison is a public service rather than a control-room screen. Choose a service such as permit handling, traffic operations or maintenance reporting. Examine who benefits, who may be excluded, and how a resident challenges an incorrect result. Evaluate the staff and procurement capacity needed after installation.

What to evaluate

Public outcome

Define service quality, safety, access, cost or policy objective.

Institutional capacity

Inspect data, staff, procurement, operations, oversight and vendor management.

Rights and risk

Assess transparency, privacy, bias, appeal, security and affected groups.

Evaluation

Set baseline, pilot, independent review, feedback and stop conditions.

Worked evaluation exercise

The following is an illustrative assessment exercise, not a reported customer result.

For a permit-processing exercise, map the steps between application and final decision. Separate document classification from eligibility decisions, identify where a human must review a case, and compare turnaround times for straightforward and difficult applications. Include an accessible non-digital route and an appeal process in the design. Ask for evidence of service quality across user groups; faster average processing alone may conceal a worse experience for applicants whose records are incomplete.

Questions for the operating team

• Which public problem is being solved?
• Can residents contest an adverse result?
• What capacity is required after procurement?

Warning signs

• Technology deployment treated as policy success.
• Affected users absent from evaluation.
• Vendor metrics replacing public outcomes.

Planning the visit

Use these questions to scope an industry-focused China AI program. Agree which records and operating workflows can be examined before confirming meetings; host participation and access require confirmation. Our research method explains how we structure the brief and follow-up.

References and scope

These references provide policy or risk-management context. They do not independently verify a host's performance or the illustrative exercise above.

• OECD AI Principles
• NIST AI Risk Management Framework