A public opinion study segmenting Californians into five "AI mindsets," so policymakers and advocates can see what moves people, who to engage, and how to build trust in AI governance.
If you have spent any time trying to build public support for serious AI governance, you already know the frustrating part: it keeps losing. Good arguments fall flat. The public seems anxious but immovable. And the people pushing for accountability are working half-blind, without a real map of who they are actually trying to reach. The California AI Compass, built with TechEquity Collaborative, was made to change that. It draws on a statewide survey and focus groups with thousands of Californians, the kind of research most advocates and campaigners would never have the budget or time to run themselves, and turns it into something you can act on. For the first time, it maps the public into five distinct AI mindsets, from Hopeful Regulators to Alarmed Populists to the Cautiously Disengaged, and shows what each one actually needs to hear.
What it reveals reframes the whole fight. 70% of Californians want strong, enforceable AI laws and reject industry self-regulation outright. 59% believe corporations and elites will reap the rewards while ordinary families are left behind, a conviction that holds across party lines. The problem was never that the public is against you. It is that no one knew how to reach them. The Compass closes that gap: it shows which arguments land with which audiences, who is genuinely persuadable, and where trust is won or lost. You can use it to sharpen a campaign message before you spend a dollar testing it, brief a legislator with evidence instead of anecdote, target the audiences most likely to move, and pressure-test your own strategy against what the public actually believes. Whatever you are working on, it makes that work land harder. Read the full interactive report, watch the overview, or request a briefing built for your sector. The effort to make AI serve the public has been losing. This is how that changes.
| Open Source: | No |
|---|---|
| Last Modified: | 2026-05-19 00:00:00 |