House Bill 1925 was introduced by Pennsylvania State Reps Arvind Venkat, Joe Hogan, Tarik Khan, Bridget Kosierowski and Greg Scott on October 6 2025. It is a bipartisan bill that would create regulations on how insurance companies, hospitals and clinicians use AI and disclose that information to patients and the general public.

 

What happened

It would also address the use of AI to the point where it may overtake the role of a human decision maker. If an insurance company, hospital or clinician were to use AI they would need to have a human make the final decision after an individualized assessment.

Insurance companies would have to attest to the PA Department of Insurance or Department of Human Services and Hospitals and Clinicians would have to attest to the Department of Health that they are taking steps to reduce bias and discrimination in their AI usage which is already illegal under PA law. They would also need to provide proof of how they are doing so.

 

What was said

Several sponsors drew on clinical experience. Venkat, the only physician in the General Assembly, said he had watched AI use in health care grow in real time, and warned it has already been shown to reinforce bias and discrimination. Scott, an EMT, said AI “should never replace the expertise or judgement of health care clinicians.” Kosierowski, a nurse of nearly 30 years, argued experienced doctors and nurses are needed more than ever to check AI findings for bias, while Hogan stressed keeping a human element in life-and-death decisions and Khan objected to insurers relying on AI rather than people to issue denials.

 

Why it matters

Having humans in the decision-making process when using AI has been backed by many studies. A scoping Health Affairs Scholar review done in 2026 about AI and health equity stated that most of the guidance they looked at stated "ethical human oversight must accompany AI-use." They also found many studies that showed AI can be harmful to those that are considered marginalized or minoritized, which is what the bill is trying to prevent with the attestation.

For there to be disclosure, there needs to be knowledge of where AI is being used. A report done by Paubox showed that 95% of healthcare organizations have staff that has access to PHI using AI in their emails, and 25% of organizations do not allow any use of AI. The report suggests having an AI usage policy and creating a BAA with your AI vendor if they will be coming into contact with PHI. With a bill like HB 1925, there could be a risk of patients' PHI being disclosed if an employee is using AI without permission.

See also: HIPAA Compliant Email: The Definitive Guide (2026 Update)

 

FAQs

What is Pennsylvania House Bill 1925?

HB 1925 is a bipartisan bill setting rules for how Pennsylvania insurers, hospitals, and clinicians use AI, covering patient disclosure, human final decisions, and bias attestations.

 

Does HB 1925 ban AI in health care?

No, the bill permits AI use but requires a human decision-maker to make the final call based on an individualized assessment.

 

Where does the bill stand now?

The House passed HB 1925 on September 30, 2026, and it has been in the Senate Institutional Sustainability and Innovation Committee since October 14, 2026.