Stanford Uses AI to Audit 3B Words of US Law, Revealing Regulatory Bloat

shashib · x · 2026-08-22

Stanford's RegLab employed AI to analyze 3 billion words of US legal code across federal, state, and municipal levels, quantifying the cost of obsolete regulations. The study found that reporting requirements in California grew 400% since 2000 with flat staffing, resulting in 30% of recurring reports never being filed. For instance, the Federal Reserve continues to report on the defunct $1 Coin Program annually because repealing mandates is legislatively difficult. This demonstrates AI's potential in cleaning up regulatory sludge and compliance auditing.

Original post →

More from Research

Research channel →