How Ai Is Helping States Cut Through Decades Of Red Tape

· Source: hai.stanford.edu · Field: Government & Public Sector — Public Policy & Governance, Regulatory & Compliance, Digital Government & E-Government · Depth: Intermediate, medium

Summary

Stanford researchers at the Institute for Human-Centered AI (HAI) and RegLab developed an AI tool to combat "policy sludge" by scanning 500 million words of state statutes across all 50 U.S. states. This tool identifies outdated reporting requirements, commissions, and fees, revealing significant bureaucratic bloat. Collaborations with New York, California, and Maryland have shown reporting requirements grew 400% in California from 2000 to 2025, with 30% of reports potentially never completed. One report alone consumed 3,500 staff hours and over \$870,000. The project provides a model state statute for automatic sunsetting and a public website to explore reporting requirements, aiming to make law legible and government programs operable.

Key takeaway

For government administrators tasked with regulatory reform, embracing AI-powered statutory analysis offers a concrete path to streamline operations. You can identify and eliminate obsolete reporting requirements, potentially saving thousands of staff hours and millions of dollars, as demonstrated in New York and California. Consider implementing model statutes with automatic sunsetting and digital tracking to prevent future policy sludge and enhance public service delivery.

Key insights

AI tools can systematically identify and help eliminate obsolete government regulations and reporting requirements.

Principles

Method

Compile state statutes, develop an AI system to scan for specific provisions like reporting requirements, then validate findings through government collaborations.

In practice

Topics

Best for: Executive, Research Scientist, Policy Maker, Legal Professional

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by hai.stanford.edu.