California publishes results of first statewide AI public deliberation project

Governor Gavin Newsom announced the results of California's first statewide public deliberation project on AI governance on October 13, 2026, releasing a report capturing resident input gathered through the state's Engaged California platform.

The report records priorities rather than binding rules, and the recommendations are meant to guide, not immediately change, state policymaking.

What it is

The report draws on Engaged California, a digital democracy platform whose AI effort was opened to every California resident in May 2026.

It is described as capturing Californians' priorities for how artificial intelligence should be developed, used and governed, built through the state's first statewide public-engagement effort.

The AI effort followed an earlier pilot tied to Los Angeles wildfire recovery, in which the platform's earlier Los Angeles firestorm recovery pilot gathered input from thousands of local residents and produced 19 recommendations for government action.

That pilot sought input from thousands of local voices to help shape recovery efforts before the tool was expanded statewide for AI policy.

What it requires

The project creates no new statute: California's first statewide public-input process on AI produced priorities for future policy, not a new compliance regime.

Three themes came up the most often and got the strongest support in the first-in-the-nation engagement.

Those were to hold AI companies accountable for the harm they cause, to protect people and the environment from the resource needs of AI, and to set rules for where and how AI can be used safely.

The strongest themes extend accountability beyond model outputs to company-caused harms and AI's infrastructure footprint, alongside rules for safe deployment.

Participants offered specific proposals, with one called for government transparency around the building, construction and implementation of data centers, and another participant suggested requiring a model to pass bias testing before public release.

Others cautioned against overregulation of industry, while one participant noted a wish for AI companies to open up their internal data to peer review.

The governor's office said the recommendations will help guide policy decisions and add to the Governor's strongest-in-the-nation AI policy framework, including nation-leading laws to protect employees in the workplace.

What security leaders should do

Because the report creates no enforceable obligations, security and compliance teams should log it as a signal of future legislative direction rather than a new control requirement to implement today. Given that data-center transparency and pre-release bias testing drew strong resident support, CISOs at AI developers and deployers with California operations should start tracking internal evidence for model evaluation and infrastructure disclosures that future bills could formalize. Teams already building governance programs around frontier-model obligations should watch whether these themes feed into amendments to existing state AI safety and transparency statutes, since officials have said the input will inform upcoming policy decisions.