A new prescription refill initiative launched in Utah has sparked a significant discussion about the potential for artificial intelligence (AI) to take on roles traditionally held by medical professionals. This program enables residents to bypass in-person visits at clinics, allowing them to refill prescriptions online through an AI chatbot named Doctronic, which aims to enhance the convenience of healthcare for both patients and doctors.
However, this development raises serious concerns among healthcare professionals, legal experts, and public health advocates. The pilot project highlights important questions about AI's integration into the medical field, such as the need for regulatory frameworks, whether physicians should have the authority to intervene in AI-driven processes, and the safety protocols necessary to protect patients.
Current regulations restrict prescription practices to licensed practitioners at the state and federal levels. Advocates for the program argue that these century-old laws need to evolve to accommodate AI technologies like chatbots. Dr. Eric Bressman from the University of Pennsylvania notes that we have reached a point where non-human entities are effectively being granted medical authority, regardless of how we define it.
While Bressman and his peers do not oppose AI involvement in prescribing, they insist that such technology must adhere to the same rigorous standards applied to human physicians, who undergo extensive years of education and evaluation before licensure.
The operational framework for Doctronic has been facilitated by a “regulatory sandbox” that allows Utah officials to bypass existing regulations for AI firms presenting innovative solutions. The administration of the refill program is managed by a panel of five AI experts, none of whom are licensed medical practitioners. They claim to have instituted various precautions, such as having human doctors review all prescriptions during the initial phase, although there are plans to move toward complete automation.
The state's medical licensing board first became aware of the Doctronic initiative through media coverage of its launch. In a letter sent to state officials, several board members expressed concern over the potential dangers of automating prescription renewals for medications that carry risks such as side effects or adverse interactions.
Dr. Alan Smith, who leads the medical board, voiced frustration at being sidelined in the decision-making process, stating they felt pushed aside regarding an initiative that could impact public safety. The issue is complicated by the division of regulatory oversight, where medical technology is typically governed at the federal level, while individual practitioners are state-regulated. Doctronic executives consider their AI system to fall within the purview of state medical practice, despite the federal Food and Drug Administration (FDA) being tasked with overseeing technologies impacting healthcare decisions—a boundary that some experts argue Doctronic has crossed.
In discussing their approach, Doctronic leaders refrained from confirming whether they have sought FDA approval, emphasizing their mission to improve patient access to healthcare. After verifying user identities, the AI gathers necessary information about prescriptions and health histories, accessing a national pharmacy database to confirm valid prescriptions. For straightforward cases, the AI completes the renewal, while more complex situations are escalated to a physician affiliated with Doctronic’s telehealth service.
Looking ahead, Oskowitz, co-founder of Doctronic, anticipates that the technology could streamline numerous routine medical functions, potentially enabling doctors to manage a substantially larger patient population.
Other states such as Texas and Wyoming are following suit, easing restrictions for AI applications in the medical field. Meanwhile, lawmakers in areas like Iowa and Idaho are exploring legislation to formally regulate AI medical services, with many proposals influenced by the Cicero Institute, a think tank that advocates for the adoption of AI in healthcare.
Opposition to medical AI mainly arises from healthcare professionals' economic concerns, noted by Cicero’s health policy director, who highlighted that initial adopters may face backlash fueled by anxiety over workforce implications and job security.
Concerns about patient safety are echoed by Smith, who warns that the list of medications eligible for automation includes high-risk drugs like blood thinners, which require careful monitoring due to potential complications. He emphasized that patient conditions can change significantly over time, meaning that previous prescriptions may no longer be appropriate.
The American Medical Association has similarly cautioned that prescription renewals are not simply administrative tasks. Officials from Utah’s AI office have indicated that the company has exercised excessive caution, often escalating non-critical decisions to physicians. To address safety concerns, several high-risk medications have been removed from the refillable list.
Utah has begun to share preliminary results from the program, and Doctronic is expected to publish peer-reviewed findings in the coming months. Currently, the only study available is an internal report that lacks independent validation, showing an 80% agreement rate between Doctronic's diagnoses and those made by human doctors in a sample of 500 telehealth consultations.
As the FDA has opted for a lenient stance on AI in medicine, some experts argue that stronger oversight should have been in place prior to Doctronic's launch. Bressman expresses concern that the current regulatory approach reflects a disconnect reminiscent of early 20th-century medical practice, underscoring the need for established standards in a rapidly evolving field.
While the FDA states its commitment to fostering medical innovation, it has not authorized any AI chatbot technologies so far. As Doctronic expands its services amid varying state regulations, experts warn that prioritizing business growth over evidence-based practices may lead to erosion of public trust and future backlash against AI technologies in healthcare.
