AI agents to run gov tasks

Gartner’s March 17, 2026 press release forecasts that at least 80% of governments will deploy AI agents to automate routine decision‑making by 2028. (gartner.com) Gartner’s conclusion is backed by a survey of 138 government respondents conducted July–September 2025 that found 41% named siloed strategies and 31% named legacy systems as key barriers to digital adoption. (gartner.com) The research calls for a pivot from model‑centric governance to decision‑centric operating models and predicts that by 2029 roughly 70% of government agencies will require explainable AI (XAI) and human‑in‑the‑loop (HITL) controls for any automated decisions that affect citizen service delivery. (gartner.com) Practical agentic use cases already being built for public IT include automated ticket triage and orchestration; vendor and engineering playbooks for AI‑powered triage workflows were published in March 2026 showing how agents can classify, extract context, and route technical tickets automatically. (orkes.io) “Living documentation” practices that auto‑generate docs from code, tests and system telemetry are being positioned as the data layer agents will query to reduce single‑person knowledge silos and keep agent context current. (kinde.com) Real‑world tension is visible in New York City, where the MyCity business chatbot was shut down in Feb 2026 after giving illegal advice while the city’s 311 system—processing more than 3.4 million requests in 2024—continues pilots for voice assistants and AI translation to handle informational demand. (newzino.com) Gartner’s earlier 2025 agentic AI forecast estimated agentic systems could autonomously resolve 80% of common customer‑service issues by 2029 and cut operational costs by roughly 30%, framing the government prediction within a broader expectation of operational savings and shifting workforce roles. (gartner.com)