ONC HTI-1 Algorithm Transparency Rule

The Health Data, Technology, and Interoperability (HTI-1) final rule updates ONC’s Health IT Certification Program. One part addresses algorithm transparency for predictive decision support interventions in certified health IT. ONC describes these requirements as providing clinical users with a consistent baseline of information to assess algorithms for fairness, appropriateness, validity, effectiveness, and safety.

The rule focuses on certified health IT and defined predictive decision support interventions. It does not create a blanket disclosure requirement for every AI product used in healthcare. Developers must identify whether a function falls within the relevant certification criteria and provide required source attributes. Health systems should understand the certified product scope and use available documentation when evaluating a tool.

Transparency supports evaluation, but disclosure alone does not show that a model performs well or is safe. Users still need to consider local population, workflow, data quality, clinical evidence, and ongoing monitoring. The rule became effective in March 2024, and organizations should check current ONC materials for implementation details and corrections. Do not confuse HTI-1 transparency requirements with FDA device authorization or other privacy and safety obligations. Implementation requires understanding whether a decision-support function is offered through certified health IT and which version is in use. A deployment may include multiple algorithms, so one disclosure should not be assumed to describe every feature. Clinicians need information in a form that supports practical review during their workflow.