Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Ai Governance topic
No spam. Unsubscribe anytime.
Little Hoover Commission proposes 'Cal Compute,' AI council and stronger procurement oversight for state government AI adoption
Summary
The Commission recommended a state-owned computing center (Cal Compute), a cross-sector California AI Council, expanded access to secure generative AI tools for state workers, standardized procurement and vendor disclosure, and strengthened worker and public engagement.
The Little Hoover Commission on Friday outlined a set of recommendations to help California state government adopt artificial intelligence tools safely and effectively.
Daniel Harris McCoy, the Commission’s project manager for the AI study, said the report builds on Governor Gavin Newsom’s 2023 executive order on generative AI and focuses on modernizing state operations rather than regulating private-sector firms. The study recommends creating a state-owned AI computing center tentatively called 'Cal Compute' (an approach noted by passage of SB 53 in related policy discussions) to provide secure, affordable computing and AI tools for state agencies, universities, nonprofits and startups.
The Commission also recommends establishing a cross-sector California AI Council to oversee AI implementation in state government, guide development of the computing center, coordinate training and procurement, and manage risk. McCoy said rank-and-file state employees should have access to secure, general-purpose generative AI tools paired with baseline training and peer-to-peer innovation supports; he cited pilot efforts such as 'Poppy AI' as a model to scale responsibly.
To reduce procurement friction while preserving safety, the report urges clearer, standardized AI-use policies across agencies, improved vendor disclosures and verification, and streamlined procurement reviews that avoid redundant assessments. The Commission recommends enhanced legislative oversight—suggesting a select committee on public-sector AI—and measures to strengthen worker trust, including worker forums and an AI ombudsperson, alongside public education about state use of AI.
McCoy also described a smaller, focused study on AI and food insecurity that found potential for AI to reduce staff time, streamline applications, improve data integration and provide multilingual services—suggesting state leadership could scale county pilots where they already exist.
What’s next: the Commission said it will continue examining procurement issues and stands ready to provide technical assistance for legislation or administrative changes implementing the AI recommendations.
