Rebooting Democracy in the Age of AI

How do you find the rules a city no longer needs? In San Francisco, AI helped, scanning 16 million words of municipal code to surface 174 outdated reporting requirements. In a new Reboot Democracy piece, Dane Gambrell walks through how the city partnered with Stanford's RegLab to draft a 351-page cleanup ordinance. Beth Simone Noveck makes the case in the Times - Scottish Edition that AI's most important democratic role may be helping governments listen at scale, drawing on Camden and Scotland. Elif Davutoglu examines Türkiye's new YAZEK system, which requires educators to formally declare every AI tool they use in classrooms, and the gap between procedural compliance and substantive oversight. And from the Code for America Summit, Robert Asaro-Angelo argues that hope in public service becomes credible when civil servants have the authority and tools to fix systems from within.

As governments search for ways to operationalize AI ethics inside public institutions, Türkiye’s new YAZEK system offers an early example of what procedural AI governance can look like in practice. The piece examines how the Ministry of Education embedded mandatory AI ethics declarations into everyday administrative workflows, while also exploring the system’s limits around verification, oversight, and remedy when automated systems cause harm.

In a new piece for the Rethinking Regulation series, Dane Gambrell examines how San Francisco used a custom AI tool developed with Stanford’s RegLab to scan 16 million words of municipal code and identify hundreds of outdated and duplicative reporting requirements across city government. The effort ultimately produced a 351-page ordinance proposing the deletion or consolidation of 174 mandates. The piece offers a grounded look at how AI can help governments make sprawling bureaucratic systems more legible and manageable.