Public bodies use AI to decide on benefits, inspections, permits and essential services, with direct effects on citizens' rights. That is why they are the first required to assess impact, keep public algorithm inventories and ensure human oversight, and they pass those requirements on to suppliers through public procurement.
Relevant sector rules
EU: access to public benefits is Annex III high-risk and public-body deployers must perform a FRIA from Dec 2, 2027. Spain: Law 40/2015 requires naming the body responsible for every automated administrative action and Law 15/2022 promotes bias assessment of public algorithms; the AI bill that would have required an AI inventory and AI officer in each entity lapsed in October 2026. Peru: the Law 31814 regulation requires public bodies to perform a prior impact assessment for high-risk uses and apply NTP-ISO/IEC 42001; the 2026-2030 national AI strategy and Directive 002-2026-PCM/SGTD require an AI Officer in each entity; executive-branch deadlines passed on Sep 10, 2026. Ecuador: the EFIA-EC strategy guides public bodies. Colombia: CONPES 4144 binds the state, not private parties. Chile: Law 21,719 covers public bodies from Dec 1, 2026. Brazil: bill PL 2338 would ban government social scoring. US: OMB memo M-25-21 governs AI use by federal agencies and Texas TRAIGA requires state agencies to disclose AI use; Colorado SB 26-189 covers essential government services from 2027.
Base documents for public bodies and their suppliers: AI lead, governance model, inventory, algorithmic and fundamental rights impact assessments, human oversight and explanation to citizens.