When a public agency uses AI, the stakeholder isn't a customer who can switch providers. It's a citizen who can't. That raises the bar on explainability, fairness, transparency and vendor accountability, and it often has to be met with small teams and tight budgets.
This session brings together perspectives from government technology leadership, audit and risk, and AI delivery. It covers what responsible AI looks like in practice: building an AI inventory, classifying use cases by risk, deciding which use cases should not proceed, disclosing AI use to residents, and governing AI that arrives embedded in vendor software. The discussion draws on NIST AI RMF, ISO 42001 and emerging state AI laws.
You'll leave with:
• A five-step starter roadmap for public-sector AI governance
• The vendor questions to ask before the next contract renewal
• A transparency checklist for citizen-facing AI