Product design question

You are launching a voice and chat agent for a state social-services program. What product requirements would you define before launch to make it trustworthy, accessible, secure, and reliable for high-stakes constituent interactions, and which requirements are launch-blocking versus iterative improvements?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Tests the ability to define launch blocking requirements for a high stakes public sector AI agent versus what can be improved iteratively.

Define trustworthiness requirements as launch blocking: accurate information, clear disclosure the user is talking to an AI, and a reliable human escalation path.
Define accessibility requirements as launch blocking, given the constituent population likely includes users with disabilities, limited literacy, or limited English proficiency.
Define security requirements as launch blocking, including data privacy for sensitive personal information like benefits or identity data.
Define reliability requirements as launch blocking, such as graceful failure handling and no false confirmations of a government action like a benefits change.
Treat response tone polish, expanded intent coverage, and personalization improvements as iterative, not launch blocking.
Set a formal go live gate combining these requirements with sign off from accessibility, security, and program stakeholders, not just product.

Separates true launch blockers, accuracy, accessibility, security, human escalation, from iterative polish, showing sound judgment about what actually protects constituents.
Flags disclosure that the user is speaking with an AI as a specific launch blocking requirement, which is often overlooked but critical in public sector trust.
Requires sign off from accessibility and security stakeholders formally, not just informally, before go live.

Treating accessibility as a nice to have improvement instead of a launch blocking requirement for a public program serving a broad population.
Launching with broad intent coverage before the reliability and escalation basics are solid, given how high the stakes are for constituents.

How would you test accessibility before launch, beyond automated compliance checks?
What would you do if a security requirement conflicts with the state's desired launch timeline?