The Evolving Landscape of AI Regulation in Medicine
As artificial intelligence (AI) continues to permeate the healthcare sector, it’s crucial to address the growing concerns surrounding its role, particularly when it starts to mimic or replace functions traditionally performed by licensed healthcare providers. Recent legislative developments across various states signal a shift towards regulating AI’s role in medicine, focusing on transparency and the safeguarding of patient care.
A New Wave of State Regulations
The burgeoning use of AI in healthcare has prompted states to take regulatory actions to clarify its boundaries within the practice of medicine. In June 2025, Nebraska became the latest state to enact legislation ensuring that prior health insurance claims undergo a final review by a qualified healthcare professional (HCP) before any denial can be issued. This move adds Nebraska to the ranks of Arizona, California, Indiana, Maryland, and North Dakota, all of which have implemented similar laws. Nearly two dozen additional states are currently contemplating akin regulations.
These laws encompass two primary components: they explicitly prevent AI from having the authority to deny or modify healthcare services without HCP oversight, and they establish rules regarding how AI may assist in this process. For instance, any AI aid must utilize individual patient data rather than broader group data to avoid discrimination against specific patient populations.
The Case of Prior Authorization Denials
The spotlight on prior authorization—a gateway through which many healthcare services are accessed—has drawn significant scrutiny. Reports indicate that about 25% of prior authorization requests face denial, leading to allegations that insurers are prioritizing profit over patient welfare. Critics argue that when decisions about medical necessity are made by AI, it raises ethical questions about who is truly practicing medicine. Courts have affirmed that prior authorization necessitates medical judgment and, as such, sanctioning AI to decide these matters could be viewed as unauthorized practice of medicine.
Interestingly, while a significant portion of denied claims (80%) are eventually overturned, the delays and administrative burdens involved can lead to deteriorating patient outcomes, highlighting the critical need for human oversight in AI-driven processes.
Transparency in AI Chatbots
The introduction of AI chatbots in healthcare has not gone unnoticed. Some states have progressed in developing regulations that demand clarity about the nature of these AI tools. For example, California has mandated the disclosure of AI chatbot utilization across all patient communications, while Colorado has taken this further by enforcing transparency in any communications involving AI for residents. Utah’s laws specifically address mental health chatbots, ensuring that users are informed that they are interacting with AI and not licensed practitioners.
Additionally, New Jersey has ground-breaking legislation in the pipeline that aims to prevent chatbots from falsely representing themselves as licensed providers of mental health services, while Oregon has banned chatbots from using terms like “nurse” to avoid misleading patients.
The Federal Landscape
These state-level initiatives are part of an evolving landscape, but they do not exist in a vacuum. The potential for federal legislation to intervene looms large. Bills like the “One Big Beautiful Bill Act,” which drew initial plans for a ten-year pause on any state regulation of AI technologies, hint at the tussle between state and federal powers in managing AI’s place in medicine. Although this specific provision did not make it into the final text of the act, it raises questions about whether a unified federal approach might be more effective than a patchwork of state laws.
Broader Implications for Healthcare Providers
As we navigate through 2025 and beyond, healthcare providers and entities must take note of the implications these regulations could have on their operations. The newly instituted state laws seem to promote a cautious approach to AI in medicine, permitting its use primarily as a supportive tool rather than a decisive authority. Such regulations advocate for maintaining the human element in healthcare decision-making, ensuring that while AI can enhance efficiency, it doesn’t compromise patient care.
Moving forward, it will be essential for businesses and healthcare providers to keep abreast of the changing legislative landscape concerning AI to mitigate risk and align with evolving regulatory requirements.
This approach to the integration of AI in healthcare not only signals the importance of transparency but also highlights the need for human oversight to protect patient rights and welfare as we enter a new era of medical technology.