Artificial intelligence (AI) has been a part of clinical research for decades, and it continues to affect how sponsors design trials, select sites, engage participants, and monitor data. However, while AI may make operations easier, there are risks. To benefit from AI in clinical research without compromising the safety and privacy of participants, sponsors must approach implementation with careful ethical evaluation and the right support.
What Do We Mean When We Discuss AI?
Variations of AI, particularly machine learning (ML), have supported trial data analysis and design for a long time. Many types of ML closely resemble advanced statistics and help identify patterns and optimize study parameters. What has changed is the rise of more complex AI, such as deep learning models. This kind of AI powers natural language processing, image recognition, generative systems, and other similar tools.
These capabilities have brought AI broader attention within clinical research. It is easy to get distracted by the small details of these discussions. However, the big picture is the same for AI adoption as it is for any technology. Researchers must protect participants, ensure transparency, and maintain compliance.
AI Benefits Require Real Use Cases
AI can deliver real value in clinical trials, but it needs to be used for the right problems. It can tailor site engagement, speed up protocol development, identify eligible participants more precisely, and automate or enhance data analysis. For example, predictive models can forecast trial bottlenecks or flag safety concerns earlier to allow quicker more informed decisions.
However, there is a caveat. Leveraging AI simply for the sake of innovation often introduces complexity without the anticipated benefits. To take advantage of AI’s power, sponsors need to use it with intention. Proper use should be aimed at a precise target, such as reducing administrative burden, improving patient matching, or enabling real-time monitoring. Tools also need to be suitably matched with their intended use.
Ethical and Operational Barriers Still Matter
Although AI can potentially streamline clinical trials, it also introduces new challenges and magnifies existing ones for human subjects protection and regulatory compliance.
Common challenges are:
- Informed consent complexity. Researchers need to educate participants about AI use and what it means for them; clear, accessible language in consent materials and early IRB collaboration can promote transparency and understanding
- AI model bias. If an AI system is trained on biased data, it will re-create and even amplify those patterns; sponsors need to prescreen datasets, implement bias checks throughout the trial, and supplement AI outputs with human expertise to enable fair, inclusive decision-making
- Privacy risks. Complex AI models can cross-reference large volumes of data. This raises the risk of identifying participants, even when data have been anonymized. Sponsors must limit access, understand the privacy implications of the models they use, and update protection when needed
- Uncertain regulatory expectations. While FDA guidance is still in draft form, sponsors can still act now. Encouraging transparency, model validation, and life cycle monitoring satisfies proposed guidance as well as aligning with IRB review principles. Adding these steps into preliminary planning can reduce friction later
Collaboration Strengthens AI-Enabled Research
As technologies become more complex, early collaboration with an IRB becomes even more important. From assessing how AI impacts participant understanding to evaluating data handling and model transparency, IRBs offer a critical layer of oversight for ethical evaluation.
At Sterling IRB, we approach AI the same way we approach every innovation: with regulatory fluency, practical insight, and a commitment to protecting participants. Our team works closely with sponsors and CROs to help navigate new tools, meet regulatory expectations, and maintain ethical integrity, no matter how technology evolves.
For a deeper look at how AI is reshaping clinical research and what ethical implementation requires, download our white paper.