Fathom Urges Rapid Implementation of California’s AI Oversight Plan

What happened
On November 18, 2026, Fathom, an AI governance group, publicly supported California Governor Gavin Newsom’s new plan for overseeing advanced AI systems. Fathom called for the state to quickly make independent risk evaluations mandatory for major AI developers. The California Government Operations Agency’s report, released the same day, recommends legislative changes to require licensed independent verification organizations (IVOs) to assess these AI systems. The plan builds on existing state laws and would put the state in charge of licensing and supervising these evaluators.
Key details
- Fathom’s CEO is Andrew Freedman.
- The GovOps report was dated November 16, 2026, and released November 18, 2026.
- The plan would require ongoing reviews of large AI developers by licensed IVOs, replacing the current voluntary system.
- IVO assessments would be submitted unredacted to the state, which would decide on any corrective action.
- The Attorney General would have authority to seek injunctions and civil penalties based on IVO findings.
Why it matters
Large AI developers in California could soon face mandatory, state-supervised risk evaluations, increasing their compliance obligations and oversight. Independent evaluators would gain access to company records and systems, with the state enforcing legal safety standards and penalties for violations. Legislators and regulators must now decide how to turn these recommendations into binding rules and ensure there are enough qualified evaluators to meet the new requirements.
