Connecticut will bar AI systems from independently denying claims or lowering provider payments for state employees starting Jan. 1, 2027. The policy affecting about 270,000 people imposes five requirements: human review of adverse AI decisions, bans on using member data to train other models, disclosure of material AI use, validation for accuracy and fairness, and oversight reporting to the Comptroller. The rule currently covers two state-administered plans; officials plan to seek a 2027 legislative expansion to all state-regulated plans.
Connecticut Bans AI-Only Insurance Denials For State Workers — Human Review Required From Jan. 1, 2027
Image credit: Suzanne Kantra/Techlicious generated by ChatGPT
Beginning Jan. 1, 2027, Connecticut will prohibit any artificial intelligence system from denying a medical claim for state employees or reducing a physician's payment unless a human reviews and approves the decision. Comptroller Sean Scanlon announced the new rule this week, using his authority over state-administered plans to restrict how insurers may use AI in coverage determinations.
What The Rule Requires
The policy applies to roughly 270,000 people enrolled in the Connecticut State Employee Health Plan and the Connecticut Partnership Plan (which covers municipal and other public-sector staff). It is structured around five core requirements:
- No AI-Only Adverse Actions: Carriers may not allow an AI tool to independently deny a claim or lower what a plan pays a provider without a human review of that adverse decision.
- Data Use Restrictions: Member information cannot be repurposed to train or develop other AI models without explicit authorization and safeguards.
- Disclosure: Insurers must disclose when artificial intelligence materially influenced a coverage determination.
- Validation: Any AI system used must be validated for accuracy and fairness before deployment.
- Oversight Reporting: Carriers must document their AI oversight procedures and report those procedures to the Comptroller's office.
"You and your doctor should make your health care decisions, not artificial intelligence," Comptroller Sean Scanlon said in the announcement.
The rule did not name specific insurers, but CT Mirror reported that Anthem administers the affected medical plans, Cigna handles dental coverage, Aetna manages Medicare Advantage for the plan, and Caremark provides pharmacy benefits. Dr. Mariam Hakim-Zargar of the Connecticut State Medical Society and Tom Swan of the Connecticut Citizen Action Group publicly supported the change.
Administratively, the rule is narrow: it covers only the two state-administered plans and does not extend to the roughly 220,000 other Connecticut residents enrolled in state-regulated commercial plans or to private insurance holders. Scanlon's office says it will ask the state legislature in the 2027 session to expand similar protections to all state-regulated health plans.
The move previews a broader national debate as insurers increasingly deploy AI to help process claims and states weigh how much human oversight they should require. By setting transparency, validation, and human-review requirements, Connecticut aims to balance efficiency gains from AI with patient protections and clinical judgment.
Help us improve.
