A Stanford research lab is using artificial intelligence (AI) to help governments clean up outdated and burdensome laws—what it calls “policy sludge”—and make public administration more efficient.
Stanford’s Regulation, Evaluation, and Governance Lab (RegLab) has unveiled the Statutory Research Assistant (STARA), a custom-built AI system that scours massive bodies of legal text to identify obsolete, duplicative, or low-value legal requirements. STARA’s first high-profile deployment came through a partnership with the San Francisco City Attorney’s Office, where it helped identify hundreds of outdated municipal reporting mandates.
“Because of the length of our code … it’s likely a project we would never have undertaken” without STARA, said San Francisco City Attorney David Chiu, who led a consultative process that resulted in a 351-page ordinance to eliminate or consolidate more than a third of the city’s 528 mandated reports.
Stanford’s brief describes “policy sludge” as a persistent and often invisible burden: old rules, overlapping mandates, and ineffective requirements that consume staff time and taxpayer dollars. The U.S. Code alone contains more than 30 million words, and previous attempts to catalog basic facts—like how many federal crimes exist—have failed due to sheer volume.
STARA tackles this by breaking down legal codes into discrete segments, applying language models fine-tuned for legal reasoning, and returning structured results tailored to real reform efforts. In San Francisco’s case, that included identifying reports on defunct programs, obsolete newspaper rack inventories, and duplicative housing data requirements.
City officials emphasized that STARA didn’t replace human oversight but dramatically reduced the cost and time of legal review. That allowed experts across departments to weigh in on which reports still serve a purpose and which could be cut.
The system’s promise extends far beyond city government. STARA has already identified over 11,000 reporting mandates in the U.S. Code, including thousands not tracked by Congress itself. Researchers say it could also help streamline advisory boards, identify outdated permitting laws, or spot legal provisions that may now be unconstitutional or discriminatory.
“For too long, City Hall has been weighed down by outdated and unnecessary code,” said San Francisco Supervisor Bilal Mahmood in support of the cleanup initiative. “This legislation will cut through that clutter.”
The researchers behind STARA say their work could serve as a blueprint for state and federal governments looking to modernize regulation and boost public sector capacity using domain-specific AI tools.
Need Help?
If you’re concerned or have questions about how to navigate the global AI regulatory landscape, don’t hesitate to reach out to BABL AI. Their Audit Experts can offer valuable insight and ensure you’re informed and compliant.
