Federal Bill Targets Corporate Influence in Health Care and PC-MSO Structures
On September 16, the Stop Corporate Takeovers of Physicians Act was introduced by Sens. Elizabeth Warren, Ron Wyden, and Jeff Merkley, along with Reps. Val Hoyle, Suhas Subramanyam, and Alexandria Ocasio-Cortez. The legislation largely mirrors Oregon’s SB 951, which took effect earlier this year, and comes amid growing state and federal efforts to address corporate involvement in health care, including increased scrutiny of physician practice management and PC-MSO arrangements in California.
Key Provisions of the proposed federal legislation include:
- Prohibition on private equity firms, insurers, and other for-profit corporations from owning or controlling medical practices.
- Closing the longstanding “friendly physician” model by restricting MSO arrangements that effectively allow investors to control physician practices despite state CPOM prohibitions.
- Bans on MSOs exercising control over core business, operational, and clinical functions, including hiring and firing decisions, physician compensation, scheduling, revenue targets, contracting, and billing practices.
- Physician owners must be licensed and actively engaged in the delivery of care within the state where the practice operates.
- Physician must retain authority over clinical judgement and a prohibition on contractual provisions, including non-compete, non-disclosure, and non-disparagement agreements are banned.
For health care investors, MSOs, physician groups, and health systems, the legislation is another signal that scrutiny of management and ownership structures is likely to intensify. Although prior efforts to implement similar legislation in the states have failed, this bill may cause more states may look to reintroduce similar legislation in 2027 legislative sessions.
For a deeper analysis of the proposal and its practical implications for providers and MSOs, see the HLB Digital Health Team’s article, Federal Bill Seeks to Outlaw the Friendly PC/MSO Model Nationwide, but State Enforcement Is the Real Near-Term Risk.
California Enacts New AI and Consumer Privacy Laws
California Gov. Gavin Newsom has signed several bills aimed at regulating the use of artificial intelligence in health care settings. Collectively, the new laws reinforce and clarify existing principles requiring meaningful human oversight of AI tools that influence clinical decision-making. Among them, AB 1979 prohibits providers from using an AI-enabled tool, system, or device to independently perform clinical functions that state law reserves for licensed professionals or from directing unlicensed personnel to perform those functions. The bill also emphasizes that licensed providers must retain independent professional judgment when evaluating and acting on AI-generated recommendations in patient care.
AB 2575 protects the ability of health care professionals to override AI-generated recommendations and prohibits developers and deployers from using a clinician’s decision to override an AI system as a defense in litigation alleging harm caused by the system’s output. SB 503 requires developers to take reasonable steps to identify, assess, and mitigate foreseeable risks of bias arising from the use of AI-assisted clinical decision support systems.
The legislature also approved, The Wellness for Oversight and Psychological Resources Act, SB 903, was signed into law and prohibits the provision or marketing of psychotherapy services through companion chatbots, including representations that a chatbot is a therapist or provides therapy. The bill, which parallels legislation first of its kind legislation which enacted in Illinois and replicated elsewhere, requires review and oversight by licensed professionals when AI is used for certain mental health-related functions, including therapeutic decision-making, triage, screening, and the detection or assessment of an individual’s mental or emotional state.
Also enacted are two notable amendments to the California Consumer Privacy Act (CCPA). AB 1542 prohibits businesses from selling or sharing a consumer’s sensitive personal information with third parties unless the consumer intentionally discloses the information or intentionally interacts with the third party. SB 923 expands the CCPA’s deletion right to cover all information a business has collected about a consumer and requires businesses that operate exclusively online to provide a dedicated online mechanism, such as a webform or portal, for consumer requests in addition to an email address. Together, these measures strengthen consumer control over personal information and expand businesses’ compliance obligations under the CCPA.
Comments on FDA’s Discussion Paper on GenAI in Health Care Due October 19
The comment period remains open for the Food and Drug Administration’s discussion paper on regulatory considerations for generative artificial intelligence (GenAI)-enabled medical devices. The paper outlines a potential framework designed to address the unique capabilities and risks associated with GenAI technologies, including a two-axis risk assessment model and a premarket evaluation approach based on “competency assessment.” Under this approach, GenAI-enabled devices would be evaluated through a combination of non-clinical benchmarking and clinical confirmation to demonstrate that they perform as intended before reaching patients. The discussion paper also explores risk-based postmarket monitoring strategies and identifies key regulatory considerations for foundation models and increasingly autonomous agentic AI systems. Collectively, the proposals offer an early look at how the FDA may approach oversight of the next generation of AI-enabled medicine. Comments are due by October 19, 2026.
Washington Court Rules that Patients Can’t Demand AI Scribe Recordings
In Raphael v. Mantei, a recently reported Washington state court decision, the judge rejected a patient’s request for access to a raw audio recording used by an ambient AI tool during a telehealth visit to assist in the creation of a clinical note. The physician had used an ambient scribe to record the encounter and generate a draft note, which was subsequently reviewed, edited, and finalized before being incorporated into the patient’s medical record. The court agreed with the treating clinic’s position that the audio recording served solely an administrative purpose and therefore fell within an exception to disclosure under Washington’s Uniform Health Care Information Act.
The clinic’s position was supported by an amicus brief filed by the American Medical Association, the Washington State Medical Association, and the Washington State Hospital Association. The groups jointly argued that AI-generated audio recordings should be treated similarly to physician dictation, handwritten notes, and other intermediate materials used in preparing the medical record, rather than as part of the medical record itself.
Although the decision is limited to an interpretation of Washington law, it highlights issues that are likely to arise as ambient documentation technologies become more prevalent across healthcare settings. The ruling does not establish a nationwide standard, address patient consent requirements, or resolve how HIPAA’s right of access applies to AI-generated recordings. Providers and health systems implementing ambient scribe technologies should continue to monitor developments in state patient-access laws while maintaining a vigilant compliance posture.
Professionals
