Federal AI Use Case Comparison

Governments are increasingly adopting Artificial Intelligence for public services. This dashboard visualizes the publicly declared AI systems used by the United States and Canada.

Administrative Context: The disclosure of AI use is still a new and evolving process for both governments.

U.S. AI disclosure fundamentally shifted between administrations. The Biden Administration (via Executive Order 13960 and OMB M-24-10) established a detailed, rights-and-safety-focused reporting framework, yielding the comprehensive 2024 dataset shown here. The Trump Administration rescinded these rules in April 2025 (via OMB M-25-21), replacing them with a highly streamlined metric focusing only on "High-Impact" systems to reduce bureaucratic friction. The consolidated 2025 U.S. inventory was published in April 2026, documenting 3,611 use cases under the new High-Impact AI framework — a nearly 70% increase in raw disclosure count over the 2,133 use cases reported under the prior rules, though direct comparison is complicated by the narrower risk focus and reduced reporting burden of M-25-21.

Canada's Prime Minister Mark Carney launched the nation's inaugural MVP AI Register in November 2025 to begin systematically tracking adoption across federal institutions. While this Register acts as a broad catalogue, Canada manages high-risk deployments through a separate Algorithmic Impact Assessment (AIA) process. The AIA mandates formal, public risk-scoring specifically for automated systems that make administrative decisions about citizens. Only the AI Register data (not the AIA data) is included here.

→ Read our record-by-record analysis of the 2025 U.S. inventory