Catalogue: AI in the Public Sector clear
10 practices not yet assessed for evidence strength.
Most practices in: United States of America (62)
Kyle, Texas built a Salesforce Agentforce-powered 311 platform, nicknamed Agent Kyle, letting residents report issues, track fixes on a map and query …
Sandia National Laboratories and the U.S. Department of Energy are testing an AI-based protective relay that can isolate electrical faults capable of …
Utah's Office of AI Policy authorized health-tech startup Doctronic to let AI review and approve routine prescription renewals under a one-year regulatory …
Virginia Governor Glenn Youngkin ordered every state agency to run Vulcan Technologies' AI system, Justinian, through the state's regulatory code, aiming to …
The FDA rolled out its generative-AI assistant Elsa agency-wide after a pilot review task reportedly dropped from days to six minutes, but …
New York's MTA retrofitted Google Pixel phones onto six subway cars to capture vibration and audio data, feeding an AI model that …
Facing a record 1M+ annual FOIA requests, DHS's Customs and Border Protection uses the RelativityOne AI platform to learn from reviewers and …
Virginia's DEQ built an AI tool that checks permit filings against checklists and cites the code section for gaps. Director Rolband reports …
CDCR has used the CSRA actuarial algorithm to score parolees' reoffending risk since 2010, yet the state told lawmakers in 2025 no …
Idaho's Dept. of Health and Welfare used a secretive spreadsheet formula, later tied to the Supports Intensity Scale, to set Medicaid budgets …
Chicago Police scored nearly 400,000 people on an algorithmic gun-violence risk list from 2012-2019. RAND's independent evaluation found no reduction in violence, …
NYC's MyCity Business chatbot, launched October 2023 on Microsoft Azure AI, was found in March 2024 to routinely give illegal advice on …
The first U.S. government task force on algorithmic transparency, created by New York City's Local Law 49, could not agree on a …
Michigan's MiDAS system auto-adjudicated unemployment fraud claims with no human review (2013-2015), wrongly accusing 34,000+ people at an ~85% error rate; a …
Open full copilot Grounded in cited practices — always check the sources.
