USArt AIAI, creative tools and media technology releases

Press Release · Fathom AI Inc.

California Issues Report Reflecting Fathom's Recommendations For Strong Independent AI Oversight

Distributed by USArt AI · Originally issued via PR Newswire

Fathom calls for speedy implementation as Newsom endorses mandatory IVO oversight for frontier AI

BERKELEY, Calif., Nov. 18, 2026 — Today, California Governor Gavin Newsom released a roadmap for implementing the state's newly-passed AI laws, reflecting many of recommendations Fathom made to the state on building out the first-ever independent verification organization (IVO) framework to oversee AI models and systems.

Fathom CEO and Co-Founder Andrew Freedman released the following statement:

"Governor Newsom's report is a detailed roadmap for how a government can quickly and effectively operationalize independent oversight of AI. It is also a full-throated endorsement of IVOs as the way to mitigate catastrophic risk.

In my opinion, this is the best a report like this can be. Conceptually strong with concrete next steps. Now, it's time to stop admiring the problem and move with all deliberate speed to execute on these solutions.

Given the rapidly advancing capabilities, this report should serve as the starting gun for action. There will be legitimate questions about how to turn this into legislative action and rulemaking. There will be debates about the power of the states and the power of the federal government. There will be disagreements over logistics, timing, and costs.

But the time for navel-gazing is over. We are eager to see the states and the federal government demand transparency, mandate that IVOs measure and address risk, and utilize traditional government powers to set meaningful guardrails on AI and prevent catastrophic harm. We will be pushing them to do so."

Fathom, the organization that designed and championed the IVO framework, had also detailed in its input to the state how IVOs can be used to address key AI governance challenges, including monitoring for recursive self improvement (RSI) and mitigating catastrophic risk. Many of Fathom's recommendations are included in the state's report:

  • Require independent evaluation for catastrophic risk for frontier labs: Under Newsom's recommendations, large frontier developments would be required to let licensed, expert, independent evaluators assess the risks their models create, moving California from a voluntary framework to a mandatory one, as Fathom recommended.
  • Hold evaluators accountable to the state, not the labs: The Government Operations Agency (GovOps) will license IVOs, oversee them continuously, and revoke the license of any that fall short. IVOs would report their findings to the state, which would decide what action to take.
  • The state sets the outcomes, and IVOs develop the methods: California would define acceptable risk in law, and IVOs would build and update the tests that measure that risk. Fathom argued that this system is what would keep oversight effective as the technology changes.
  • A clear legal test: The report includes Fathom's recommendation that a frontier model should be "unlikely to increase the risk of catastrophic harm above the level that would exist absent the model," and that its benefits should outweigh its remaining risk, closely mirroring language Fathom suggested.
  • Prioritize evaluating for Recursive Self Improvement (RSI): Fathom urged California to prioritize recursive self-improvement, and the report recommends making the assessment of automated AI research and development the first task for IVOs.
  • Grow the market for IVOs now: Fathom argued that California should not wait for the evaluator market to mature before imposing requirements on developers, noting that a mandate is critical to drawing the necessary capital and talent to build out the IVO sector. The report reflects Fathom's suggested licensing requirements phasing in as it grows.

California passed the first-ever IVO framework into law this summer via SB 813 and AB 1405, with Governor Newsom accelerating its implementation in a subsequent executive order, and naming Fathom Senior Advisor Dr. Gillian Hadfield to his panel of AI experts.

About Fathom: Fathom is an independent nonprofit whose mission is to build a governance architecture that helps society navigate the transition to a world with AI by fostering trust, safety, and innovation. Fathom has developed and championed the independent verification model for AI and works with policymakers across the country to put it into practice. Learn more at http://fathom.org.

SOURCE Fathom AI Inc.