In a decisive move that signals a significant philosophical shift in federal oversight, the Trump administration has begun systematically unwinding key health information technology regulations established during the previous administration. The Department of Health and Human Services (HHS), through its Office of the National Coordinator for Health Information Technology (ONC), introduced two proposed rules in late 2025 designed to drastically reduce regulatory burdens, enhance clarity for developers, and lower operational costs within the health technology sector. This pivot towards deregulation is particularly pronounced in the rapidly evolving field of artificial intelligence, where new policies are set to replace transparency mandates with a framework that prioritizes innovation and reduces compliance overhead. This initiative represents one of the most substantial reversals of health IT policy in recent years, prompting a reevaluation of the balance between regulatory safeguards and technological advancement across the healthcare industry.
A New Direction for AI Transparency
A central pillar of this new deregulatory push is the proposed elimination of the AI “model card” requirement, a signature policy of the Biden era that only took effect at the beginning of 2025. This regulation had mandated that health IT vendors create a type of “nutrition label” for their AI-powered clinical decision support tools, compelling them to provide detailed transparency reports. These model cards were intended to demystify the inner workings of complex algorithms by requiring companies to disclose critical information about how their AI models were designed, the specific datasets used for training, and the ongoing processes for maintenance and performance evaluation. The primary goal was to mitigate the inherent risks of biased or faulty AI outputs that could lead to adverse patient outcomes, giving providers and health systems a clearer understanding of the tools they were implementing in clinical settings. This approach was heralded as a foundational step toward building trust and accountability in healthcare AI.
However, the Trump administration contends that these transparency requirements have failed to deliver on their promise. In the proposed rule, designated as HTI-5, the ONC asserts that there is no compelling public evidence to suggest that the model card mandate has resulted in tangible positive clinical outcomes, such as the identification and removal of deficient or biased algorithms from the market. The rule states that the anticipated savings and patient safety benefits of this heightened transparency may not have been realized in practice. From the administration’s perspective, the regulation has instead placed a significant and costly burden on technology developers without a demonstrable return on investment for the healthcare system. Consequently, the proposal to rescind the requirement reflects a belief that market forces and existing professional standards are sufficient to ensure the quality and safety of AI tools, and that prescriptive federal mandates can stifle the very innovation they are meant to guide.
Overhauling Health IT Certification and Data Exchange
Beyond the realm of artificial intelligence, the HTI-5 rule proposes an aggressive overhaul of the ONC’s foundational health IT certification program, a critical pathway for software vendors seeking to have their products qualify for federal incentives. The proposed changes are sweeping, with the ONC planning to remove 34 of the 60 existing certification criteria and substantially revise another seven. This action would effectively revamp nearly 70% of the entire program, marking a profound shift in how the federal government evaluates and validates health technology. The administration’s rationale is that many of the current requirements have become obsolete or redundant. It argues that numerous criteria now reflect well-established industry best practices that vendors would follow regardless of regulation, while others are duplicative of rules enforced by other agencies or simply place an excessive and unnecessary compliance burden on technology companies, particularly smaller innovators who may lack the resources of larger corporations.
In a parallel but equally significant action, the Department of Health and Human Services is formally withdrawing several outstanding proposals from a comprehensive 2024 rule issued by the Biden administration. That rule was specifically crafted to improve data sharing among insurers, providers, and public health organizations, aiming to break down the persistent information silos that were widely seen as hampering the national response to the COVID-19 pandemic. The provisions now being rescinded include new certification standards for certain application programming interfaces (APIs) designed to facilitate seamless data flow, as well as a first-of-its-kind requirement for health IT vendors to enable the exchange of standardized public health data with government agencies. The withdrawal of these measures signals a retreat from federally mandated interoperability solutions, favoring a more market-driven approach to health data exchange and integration.
The Broader Implications and Next Steps
These targeted deregulatory actions in health IT are not isolated events but are instead consistent with a much broader administrative agenda. The moves directly align with an executive order President Trump signed on January 31, which directed all federal agencies to identify and reduce regulatory burdens across the economy. Following this directive, HHS has publicly signaled its intent to “dramatically expand its deregulatory efforts,” a commitment further demonstrated by a recent request for information seeking public input on how to accelerate the adoption of artificial intelligence in healthcare. While the healthcare industry has shown considerable eagerness to leverage AI for streamlining administrative tasks such as billing and scheduling, a more cautious sentiment prevails regarding its application in direct patient care. The high stakes involved in clinical decision-making mean that providers and health systems remain wary of adopting AI tools without robust evidence of their safety, efficacy, and fairness.
The regulatory landscape for health technology has now entered a period of transition and feedback. The comprehensive HTI-5 rule, which includes the elimination of AI model cards and the overhaul of the certification program, has been opened for a 60-day public comment period, allowing stakeholders from across the healthcare ecosystem to provide their perspectives. In stark contrast, the withdrawal of the Biden-era data-sharing and public health provisions was designed for a more immediate impact, as it was set to become effective as soon as it was published in the Federal Register, which was scheduled for December 29. These developments left health IT developers, healthcare providers, and patient advocates to navigate a rapidly changing set of federal expectations, reshaping the future of digital health in the United States.
