U.S. Legislation Proposes AI Systems as Legal Drug Prescribers
核心洞察
The Healthy Technology Act of 2025, introduced by Rep. David Schweikert, seeks to classify AI models as qualified health practitioners with prescription authority.
The groundbreaking legislation would require AI prescribers to receive both state authorization and FDA approval, though specific implementation details remain undefined.
Medical experts express caution, citing a recent Lancet Digital Health study that highlights both potential benefits and risks of AI in clinical decision-making.
A transformative legislative proposal in the United States could fundamentally alter the landscape of medical prescribing practices by enabling artificial intelligence systems to prescribe FDA-approved medications. Representative David Schweikert (R-AZ) has introduced the Healthy Technology Act of 2025 (HR 238), which aims to expand the definition of healthcare practitioners to include AI and machine learning technologies.
Proposed Legislative Framework
The bill, currently under review by the House Committee on Energy and Commerce (搜索), would amend the Federal Food, Drug, and Cosmetic Act to recognize AI systems as legitimate prescribers. To qualify, these AI systems would need to secure both state-level authorization and FDA approval. However, the notably brief legislation leaves crucial details undefined, including specific drug categories and clinical scenarios where AI prescribing would be permitted.
Current AI Applications in Healthcare
While the proposal may seem premature to some, AI is already making significant inroads in healthcare delivery. Large language models (LLMs) currently assist medical specialists, particularly radiologists, in diagnostic processes that inform prescribing decisions. These systems also streamline routine administrative and clinical tasks, demonstrating their growing utility in healthcare settings.
Expert Concerns and Scientific Evidence
Medical professionals have expressed skepticism about the immediate feasibility of AI-driven prescribing. Their concerns center on the technology's current limitations in reliably determining medication safety and efficacy for individual patients.
Recent research published in The Lancet Digital Health (搜索) offers a measured perspective on AI's healthcare applications. While acknowledging potential benefits such as reduced physician workload and improved decision consistency, the study emphasized that existing LLMs require further evaluation before being entrusted with clinical decision-making responsibilities.
Parallel Legislative Initiatives
Alongside the AI prescribing bill, Rep. Schweikert has introduced complementary legislation (HR 193) directing the Health and Human Services (搜索) department to develop Medicare payment guidelines for AI-enabled health technologies, particularly in remote monitoring applications. This multi-pronged approach suggests a broader strategic push toward AI integration in healthcare delivery systems.
Implementation Challenges
The path to AI prescribing faces numerous hurdles, including technical limitations, regulatory frameworks, and professional acceptance. Healthcare institutions must carefully balance innovation with patient safety, particularly given the complex nature of clinical decision-making and the potential risks of automated prescribing systems.
Medical experts emphasize that while AI shows promise in supporting healthcare delivery, maintaining high standards of patient care quality remains paramount. The successful implementation of AI prescribing systems will require robust validation processes and careful consideration of both benefits and potential risks.