> ## Content Index
> Fetch the complete content index at: https://www.fdaweb.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Uses of and Issues with AI in Trial Design
- URL: https://www.fdaweb.com/uses-of-and-issues-with-ai-in-trial-design/
- Published: 2024-05-30T12:00:00.000Z
- Updated: 2026-09-14T14:32:05.000Z
- Author: David McFarland
- Tags: Drugs, #legacy-id-D5157089

CDER Office of Medical Policy director **Khair ElZarrad** says the use of artificial intelligence (AI) in drug regulatory submissions has seen a rapid increase in recent years. Interviewed on a *Q&A with FDA* [podcast](https://www.fda.gov/drugs/news-events-human-drugs/role-artificial-intelligence-clinical-trial-design-and-research-dr-elzarrad?ref=fdaweb.com), ElZarrad said that from 2016 to 2024, some 300 submissions were received that reference AI use.

“These submissions \[traverse\] the landscape of drug development, all the way from discovery to clinical research, and also to post-market safety surveillance and even manufacturing,” he said.

ElZarrad said AI is being used to analyze a vast amount of data from both clinical trials and observational studies. “Its use is really to help make inferences regarding the safety and effectiveness of the drug being evaluated,” he said. “AI also has the potential to inform the design and efficiency of clinical trials, including in decentralized clinical trials. And you also see a potential for AI use in trials incorporating the use of real-world data.”

Asked about other potential applications, ElZarrad mentioned the use of AI in predictive modeling, improving clinical trial conduct, and assisting in recruitment.

Concerning drug safety, he said AI-enabled algorithms can detect clusters of signs and symptoms to identify potential safety signals and can do that in real time. “AI can be used to predict also adverse events in clinical trial participants,” he said. “And this is an area we are definitely interested in exploring.”

Unique challenges in using AI, ElZarrad said, include the variability in the quality, size, and representativeness of data sets for training AI models, the difficulty in understanding how AI models are developed and how they arrive at their conclusions, and the fact that AI model performance could degrade over time.