Strategic Report  ·  2026-10-01

Implementing the Transparency in Frontier AI Act Report for the California Department of Technology Multistakeholder Workshops

Strategic ReportMedium impactUnited States
Rob Reich (Stanford HAI Associate Director) and Deirdre Mulligan (UC Berkeley School of Information) synthesized two August 12–13 multistakeholder workshops — hosted by Stanford HAI, UC Berkeley, CITRIS, CLTC and the Berkeley Risk & Security Lab — into concrete recommendations for the California Department of Technology's first annual report under the Transparency in Frontier AI Act (SB 53). The paper draws on input from researchers, civil society, industry and state officials to advise CDT on implementing the law's frontier-AI transparency requirements; the recommendations are the authors' alone though they reflect the workshop discussion.
California's TFAIA implementation is the template other frontier-AI statutes (New York RAISE, Colorado, Texas) are watching — the paper signals the disclosure-and-reporting duties that will land on frontier developers, and the annual-report lens regulators are adopting.
Read it to anticipate the transparency/safety-reporting requirements California will formalize for frontier model developers, especially if your organization operates frontier systems in the state.
Stanford HAI — Implementing the Transparency in Frontier AI Act Report
See this in the live feed Explore related AI security and governance findings — updated every morning.
Open the feed →