AI-based tools to support government decisionmaking, implementation, and interaction are already at work in many federal agencies. Yet little is known about how agencies acquire, use, and oversee such AI systems. In an effort to fill these gaps, the Administrative Conference of the United States commissioned a report from researchers at Stanford University and New York University entitled Government by Algorithm: Artificial Intelligence in Federal Administrative Agencies.
From the report:
Table 1 previews some of the use cases explored in this report and advances a typology of governance tasks to which agencies are applying AI. Among these are two core tasks of modern government: enforcing regulatory mandates (“enforcement”) and adjudicating benefits and privileges (“adjudication”). However, federal-level use cases span well beyond enforcement and adjudication to other critically important governance tasks, such as regulatory analysis, rulemaking, internal personnel management, citizen engagement, and service delivery.
Hat tip to beSpacific