Fathom Applauds Connecticut Passing Legislation that Establishes Independent Verification (IVO) for AI Governance
CT General Assembly overwhelmingly passes bipartisan legislation establishing the Independent Verification Organization (IVO) framework; Connecticut becomes second U.S. state to advance the IVO model
HARTFORD, Conn., May 7, 2026 /PRNewswire/ -- Yesterday, the Connecticut General Assembly passed House Bill 5222, legislation that includes a pilot program for independent verification of AI systems in Connecticut. The Independent Verification Organization (IVO) provision of the bill, championed by Senator James Maroney, Chair of the General Law Committee, establishes a voluntary framework through which expert-led IVOs assess whether AI products meet safety standards set by the state. HB 5222 — along with an earlier bill that included similar language (SB5) — passed both chambers with strong bipartisan support, reflecting growing momentum behind independent verification as a pro-innovation, pro-safety alternative to traditional "command and control" regulation or industry self-policing.
The IVO provisions within HB 5222 establish a governance approach designed to increase safety, trust, and innovation by giving consumers and businesses confidence in the AI products and systems they use. Under the framework, the state sets outcome-based safety goals and authorizes a marketplace of independent verification organizations to develop technical criteria and assess whether AI products meet those goals. AI companies voluntarily elect to have their products verified, and those that achieve verification earn a trusted seal of approval and evidentiary support in the event of litigation. The approach is modeled on governance structures that have proven effective in other industries, such as financial auditing and product safety certification, and is designed to keep pace with the speed of AI innovation while solving for the technical deficit governments face in overseeing these systems.
"Connecticut is focused on being a leader in responsible AI," said Senate General Law Committee Chairman James Maroney. "One of the biggest challenges with AI policy is finding the balance between meaningful safeguards and continued innovation, while recognizing that the technology changes faster than the ability to regulate it. The IVO framework provides a flexible framework that allows us to do both. This approach incentivizes having independent experts evaluate AI systems against clear safety standards. It is an approach that we believe keeps pace with the rapidly evolving technology while ensuring responsible development."
"This pilot program reflects what Connecticut does well — bringing people together to tackle complicated issues in a practical, bipartisan way," said House General Law Committee Ranking Member David Rutigliano. "We approached this with a light regulatory touch, but with a clear understanding that the public trust matters. Residents deserve confidence that AI systems are being independently evaluated for safety, and businesses deserve a framework that is predictable, workable, and encourages responsible innovation."
"This week, Connecticut made history. By authorizing this IVO pilot program, the Connecticut General Assembly shows that safety and innovation aren't competing values but reinforcing ones," said Fathom Co-Founder and CEO Andrew Freedman. "This is the proof of concept the country has been waiting for: a governance model that empowers independent experts, earns public trust, and gives industry the certainty it needs to build responsibly. Connecticut lawmakers, led by Sen. James Maroney, are demonstrating that we don't have to choose between moving fast and moving safely — and every other state should be paying attention."
"This is what smart AI governance looks like in practice," said Dr. Gillian Hadfield, Bloomberg Distinguished Professor of AI Alignment and Governance at Johns Hopkins University. "This legislation takes the concept of independent verification — government sets the safety outcomes, licensed experts verify whether those outcomes are met, and companies opt for the benefits of being verified by independent third parties — and turns it into a working framework. The architecture for trustworthy AI is no longer theoretical. It's being built, starting in Connecticut, right now."
In passing this legislation, Connecticut became the second state in the nation to advance the IVO model through legislation and the first to fully authorize and implement a pilot program. Last month, Virginia Governor Abigail Spanberger signed SB 384 / HB 797, tasking the state's Technology and Science Commission to study IVOs. Under Section 47 of HB 5222, the Department of Consumer Protection will administer a multi-year pilot program approving up to five IVOs to verify AI models against risk mitigation and harm prevention standards; AI companies that achieve verification gain evidentiary support in private civil litigation, creating a meaningful market incentive for voluntary participation. The pilot includes a built-in evaluation by UConn's Institute for Municipal and Regional Policy, with recommendations on whether to make the program permanent and establish reciprocity with other states. This legislation earlier passed as SB5, which also passed with overwhelming support: 32-4 in the Senate, 131-17 in the House.
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. Learn more at http://fathom.org.
SOURCE Fathom AI Inc.
Share this article