New Clinical Studies Involving AI Registered (Monthly)

## What this measures

The number of clinical studies of any kind (trials and observational studies) newly registered on ClinicalTrials.gov each month whose record mentions artificial intelligence, machine learning, deep learning, large language models, ChatGPT or generative AI.

## Why it matters

Medicine is one of the places AI is being tested on people, in diagnosis, imaging, triage and decision support. Registered studies show how fast that research is growing and when it turns from models on old data to studies in patients.

## How to read it

Higher means more new studies involving AI. From a handful a year before 2016 it passed a thousand a year in 2024. Read with the AI share (clinicaltrials.registered.ai_share), which allows for the growth of all registrations.

## What it does not say

A registration is a plan, not a result: some registered trials never start, and size, cost and odds of success differ widely. Being listed does not mean a study was reviewed by the U.S. government.

## How it's built

Records matched by ClinicalTrials.gov's own full-text search for these phrases, in any field of the record: “artificial intelligence” OR “machine learning” OR “deep learning” OR “large language model” OR “large language models” OR ChatGPT OR “generative AI”. A study counts once however many phrases it has. Search covers the current text of each record, so a study that adds AI wording later moves into its registration month. Each record is dated by the month ClinicalTrials.gov first posted it, a date that never changes. Registration is the timely measure of new trials: U.S. law and journal editors require a trial to be registered by the time it starts enrolling, and most register before. The trial's actual start date is filled in later, over six to twelve months, so a count by start month undercounts every recent month. Records are read in full from the API each week; the last three months are rewritten. History from January 2010.

## Source

ClinicalTrials.gov (U.S. National Library of Medicine), API v2, refreshed weekly by /api/cron/clinicaltrials-pipeline-sync. Palanor counts and classifies the records as described above; it does not change them. ClinicalTrials.gov disclaimer: the U.S. government does not review or approve the safety and science of all studies listed; sponsors and investigators are responsible for their listings.