Content Highlights
The hospital expects you to use AI, but you question whether its recommendations are right for your patients. Could protecting your clinical judgment put your job at risk? Before you agree or refuse, you need to understand your rights.
At Gardner Employment Law, we can help you review your employment agreement and evaluate your options when an AI requirement threatens your professional independence.
Does the AMA Support Your Right to Question AI?
Yes. The AMA supports physician oversight of patients and opposes penalizing physicians for declining the use of AI during the current period when key safeguards and standards remain in flux. However, that position is professional policy, not an automatic legal exemption from your employer’s requirements.
In its June 10, 2026 announcement, the AMA adopted policies emphasizing that AI should assist clinical decisions while preserving physician judgment. The policies call for standards addressing evidence, validation, transparency, and explainability. It also opposes replacing physician review with autonomous or semiautonomous AI in insurance coverage determinations.
A separate AMA policy, Augmented Intelligence in Health Care, addresses penalties head on. The AMA specifically opposes penalizing physicians who decline using AI because of the unsettled status of regulatory oversight, clinical validation, usefulness, and standards of care. There is a qualification to that statement: the policy does not endorse refusing every AI system indefinitely.
Can your employer still discipline you? Potentially, yes. As the Texas Workforce Commission explains, Texas generally follows the at-will employment rule unless a contract or law provides otherwise. A hospital generally is your employer. As such, the hospital can establish workplace requirements that you are obligated to follow. Refusing to adhere to a hospital policy can put your employment at risk. Whether a particular penalty is lawful requires examining your employment agreement and applicable legal protections.
Here is a suggested analytical procedure. First, identify the actual problem. Are you refusing training required by the hospital? Are you declining to use a documentation tool, maybe a new electronic record platform? Or – are you rejecting a clinical recommendation because it does not fit your patient’s condition? Each situation that you face deserves separate analysis. Only when the hospital’s requirement affects patient care or safety will you have standing to question or even refuse to follow the policy, as a general rule.
For example, imagine AI recommends discharging your patient, but you identify findings that warrant continued observation. Ask management how you can document your assessment and depart from the recommendation before a disagreement becomes a disciplinary issue.
What Have Courts Said About AI Replacing Physician Judgment?
A federal court has allowed contract claims involving alleged AI substitution for physician review to proceed. As explained below, that ruling does not establish your right to refuse AI across the board.
In Estate of Gene B. Lokken v. UnitedHealth Group, Medicare Advantage beneficiaries alleged that UnitedHealth used an AI model (called “nH Predict”) to deny post-acute care coverage despite treating physicians’ recommendations. They also alleged that the insurance company required its staff to follow the model’s criteria under threat of termination. The lawsuit is ongoing, but the district court’s rulings are relevant to how courts view statements made by an employer about its operations.
The insurance company asked the court to dismiss the case. But the Minnesota federal district court denied the motion and allowed claims for breach of contract and breach of the implied covenant of good faith and fair dealing to proceed.
The beneficiaries offered evidence which included documents created by the insurance company that promised decisions involving clinical staff and physicians. This evidence prompted the judge in a later discovery order to require production of certain performance, compensation, and disciplinary records contained in the insurance company’s files pertaining to post-acute care coordinators, including medical directors employed by the insurance company.
This is a coverage case rather than a malpractice decision about an incorrect AI diagnosis. Its practical relevance to you is the importance of written promises about physician involvement.
If management promises that you retain clinical discretion, during negotiations request specific language explaining how that discretion operates when you disagree with AI. Who reviews your objection? Can you override the recommendation? Will doing so affect your evaluation or compensation?
In your current contract, look for wording that gives you the right to make decisions that affect patient care or safety. The wording could be used to argue that AI cannot be used instead of your clinical judgment based on your treatment of the patient.
How Can You Protect Yourself Before Refusing AI?
Review your contract, document your concerns, and request safeguards before refusing an AI requirement. A specific explanation helps your employer address the problem and gives your attorney a record to evaluate.
Check your employment agreement. Review provisions requiring compliance with workplace policies and designated technology. Examine the definition of “Cause,” any protections for independent clinical judgment, and whether you receive notice and an opportunity to correct an alleged violation before termination.
Clarify what your employer requires. Must you use the tool, review its output, or follow its recommendations? Ask how you can override an error, protect patient information and patient safety, and report concerns without affecting your evaluation or compensation.
Document the specific problem. Identify inaccurate output by the AI tool, missing information, or a patient-specific reason for disagreeing. Use approved reporting channels and keep patient information within authorized systems.
Request a workable alternative. Additional training, manual documentation, or physician review may resolve the disagreement.
These steps help you prepare your response. This process do not determine whether refusing AI is legally protected, but this may help you collaborate with management to find a workable solution.
Contact Us Today.
If management threatens discipline, seek advice before resigning or assuming your objection is legally protected. At Gardner Employment Law, we are glad to review your agreement and evaluate your options.