Who Loses Their Nursing License When AI Gets It Wrong?
Artificial intelligence is rapidly changing healthcare. From electronic medical records that automatically generate nursing notes to algorithms that predict patient deterioration, AI promises to make healthcare more efficient, reduce errors, and give nurses more time to spend with their patients.
But what happens when AI gets it wrong? More importantly, who is responsible when a nurse relies on AI-generated information that results in patient harm?
As a nurse attorney who represents nurses before their licensing boards, I believe this is one of the most important questions facing our profession today. Technology may be changing how nurses practice, but one thing has not changed: Nurses remain accountable for their own professional judgment.
Whether nurses realize it or not, artificial intelligence is already influencing many aspects of patient care. Hospitals are implementing AI-powered systems that analyze vital signs, identify patients at risk for sepsis, predict falls, assist with medication management, and generate clinical documentation. Some nurses are also turning to generative AI tools to help draft patient education materials, summarize medical information, and formulate nursing care plans.
There is tremendous potential here. AI may help nurses recognize subtle changes in a patient’s condition, reduce administrative burdens, and improve clinical decision-making. However, AI is not infallible. These systems can generate inaccurate information, overlook important clinical findings, or make recommendations based on incomplete or biased data. Generative AI can even produce information that sounds entirely credible but is completely fabricated. This is called a hallucination. Unlike a human colleague, AI does not necessarily recognize when it is wrong.
Imagine that you are caring for a postoperative patient whose vital signs are beginning to change. The patient’s heart rate is increasing, blood pressure is trending downward, and the patient seems unusually restless. However, the hospital’s AI-powered early warning system indicates that the patient is at low risk for deterioration. Do you trust the technology, or do you trust your nursing assessment?
Now imagine that the patient subsequently experiences a serious complication. If the nurse failed to recognize and respond to the patient’s deteriorating condition because the AI system indicated there was no immediate concern, the nurse’s clinical judgment
could come under scrutiny. The hospital may have purchased the technology, and the software company may have developed the algorithm, but the nurse is still responsible for exercising appropriate nursing judgment. Following an AI recommendation does not automatically protect a nurse from allegations of negligence or professional misconduct.
Another area of concern involves AI-generated nursing documentation. Imagine using an AI tool to generate a nursing note. The note sounds professional, includes appropriate terminology, and appears to capture the patient’s condition. There is just one problem. The AI documents that the patient was educated about fall precautions, but you never provided that education. Or perhaps it states that you reassessed the patient’s pain after administering medication, even though that reassessment never occurred. You review the note quickly, sign it, and move on to your next patient.
Months later, a complaint is filed with the nursing board. The medical record now contains documentation of nursing interventions that were never performed. Who is responsible? Potentially, you are. Signing an AI-generated nursing note generally means you are attesting to the accuracy of that documentation. Inaccurate documentation can raise concerns about falsification of records, failure to meet professional standards, and the credibility of the nurse. Even when an error is unintentional, it can create serious problems during a licensing board investigation. Remember, AI does not have a nursing license to lose. You do.
Patient privacy presents another significant risk. Suppose a nurse wants help writing a patient education handout or summarizing a complicated clinical situation. The nurse copies information from the patient’s medical record and pastes it into a publicly available AI application. Depending on the information disclosed, the platform used, and the safeguards in place, this could create a HIPAA violation or violate the employer’s confidentiality policies. Removing the patient’s name does not necessarily make the information anonymous. Dates, diagnoses, locations, and other details may still make a patient identifiable. Nurses must understand that not every AI platform is approved for handling protected health information.
Can nurses actually be disciplined for following AI recommendations? Potentially, yes. Although nursing boards are still confronting the implications of artificial intelligence, existing nurse practice acts already establish expectations regarding professional competence, clinical judgment, documentation, and patient safety. A board does not necessarily need an AI-specific regulation to investigate conduct involving inappropriate reliance on technology.
For example, if a nurse administers medication based on an AI-generated recommendation without verifying the order or identifying an obvious
contraindication, the board may examine whether the nurse complied with applicable medication administration standards. Similarly, if a nurse fails to intervene because an AI monitoring system did not generate an alert, the board may evaluate whether the nurse met the standard of care. Of course, liability is not automatic. The circumstances matter, including the nurse’s role, available information, employer policies, training, and whether the nurse acted reasonably under the circumstances. There may also be questions about the responsibility of the hospital, software developer, or other healthcare professionals.
I believe one of the greatest risks of AI is not necessarily that it will make mistakes. It is that nurses may become so accustomed to relying on technology that they begin to question their own clinical judgment. We have already seen how electronic health records can encourage nurses to focus more on completing documentation than observing the patient in front of them. AI could take that problem one step further.
There is even a name for this phenomenon: automation bias. It occurs when people place excessive trust in automated systems, sometimes overlooking contradictory information. When technology consistently provides recommendations, it becomes tempting to accept those recommendations without critically evaluating them. For nurses, this can be dangerous. Your assessment skills, critical thinking, and ability to recognize subtle changes in a patient’s condition are essential components of safe nursing practice. AI can support those abilities, but it should never replace them.
I am certainly not suggesting that nurses avoid artificial intelligence. In fact, AI is here to stay and getting more powerful each day. However, nurses need to approach this technology with the same professional skepticism they would apply to any other clinical tool. Before relying on AI-generated information, ask yourself whether the recommendation makes sense in light of your own assessment and the patient’s clinical presentation. Verify important clinical information through appropriate sources, and never assume that an AI-generated nursing note accurately reflects the care you provided.
If an AI recommendation conflicts with your assessment, investigate the discrepancy, communicate your concerns, and escalate the situation when necessary. Document your clinical findings, interventions, communications, and rationale accurately. If a complaint is ever filed, your documentation may become one of the most important pieces of evidence in your defense.
There is also an important question that healthcare organizations need to address. If a hospital requires nurses to use an AI system, should the nurse bear the concequence?
Comments are closed