Public Sector AI Governance

EU AI Act (Annex III) lists migration, asylum, border control, and law enforcement AI as high-risk - government use faces the strictest scrutiny
Art. 5 Prohibitions explicitly ban real-time biometric identification in public spaces and social scoring by government bodies
US Executive Orders on AI establish governance requirements for federal agency AI use, including impact assessments and public transparency
State & local procurement requirements increasingly mandate AI governance evidence as a prerequisite for government contracts

Many government AI uses are explicitly high-risk or prohibited under the EU AI Act - benefits adjudication, law enforcement, and immigration all face Annex III requirements
Real-time biometric identification in public spaces is generally prohibited (Art. 5) - violations carry fines up to EUR 35M or 7% of global revenue
Social scoring by government bodies is explicitly prohibited - AI that evaluates citizens' trustworthiness based on social behavior is banned
Transparency requirements demand citizens be informed when AI is used in decisions affecting them
Procurement requirements increasingly mandate AI governance evidence, turning compliance into a market access requirement

Instant detection of prohibited AI practices (Art. 5) before deployment --- prevents catastrophic compliance violations

Autonomy Boundary Guard prevents AI from exceeding its authorized scope --- enforces strict operational limits on government AI

RACI matrix ensures accountability across government departments --- maps ownership from deployment through operations

Audit trail provides complete evidence chain for public accountability and FOIA compliance requirements

Citizen services, law enforcement, and government AI accountability