Human In The Loop (HITL) Still Required for Clinical Trial Data Collection/Processing

Newsletter Date: December 3, 2025

With the adoption of LLMs and Generative AI in the clinical trial process it is important to remember that current government regulations require humans (medical professionals) to take responsibility for all data collected and submitted for product approvals.

The advances in GenAI and LLMs are giving sponsors the ability to collect subject data from unstructured text, voice, and images which make up a large percentage of patient/subject data. This allows for a more complete “picture” of a subject’s reaction to a new drug, device, or therapy.

While the application of new technology for improving the clinical trial data collection process is very real, the need for human oversight still exists and must continue. This requires Human In The Loop (HITL)-enabled application development, where human approval of all clinical trial data is built-in to the process prior to entry into the trial database. 

We need to think of Generative AI as “suggested” input rather than absolute truth. In this context it is a powerful tool which can be used by sponsors and investigative partners to improve the process, quality and expediency of collecting, managing, and analyzing clinical trial data.