´óÏóÊÓƵ

In recent days, we at ´óÏóÊÓƵhave been excited by the enthusiastic response to a  that invites grant proposals to examine the impact of artificial intelligence (AI)* on patient safety. 

´óÏóÊÓƵfunding opportunities aim to address today’s essential challenges in advancing and improving healthcare delivery. However, this one may carry particular weight because we believe that effective, safe AI use has the potential to profoundly change American healthcare.

AI is already revolutionizing nearly every aspect of our lives, including healthcare. This seems obvious, yet the impact of AI in our field cannot be overstated. As healthcare enters its third era in the use of AI,  transformations are occurring in diagnosis, treatment, and disease surveillance. Generative AI holds an even more tremendous promise to transform care delivery.

This funding opportunity coincides with the 20th anniversary of AHRQ’s , which has played a central role in advancing the knowledge base in technologies like clinical decision support (CDS) and telehealthcare. From the beginning, the Digital Healthcare Research Program has anticipated where the field is going, supporting groundbreaking research and demonstration projects that continue to guide U.S. healthcare. 

This remains true with AI. ´óÏóÊÓƵhas funded or is currently funding  with an AI component. These comprise approximately  and cover many technologies and aspects of care.

For example, ´óÏóÊÓƵrecently awarded a  to alleviate hospital emergency department overcrowding by developing a CDS tool that uses machine learning algorithms to improve triage. Researchers believe that using this tool will improve the prediction of patient acuity and complexity, leading to safer, higher-quality, and more equitable care. Other Digital Healthcare Research Program projects that contain an AI component treat aspects of care, including , , and . 

These all hold great promise. But a note of caution. We face four immediate challenges as we apply evidence to . 

First, we must reevaluate our research approach. Even as we employ existing frameworks, an AI-enabled world demands a more agile approach. Traditional grant mechanisms that fund long studies remain essential, but will not produce evidence fast enough to keep up with demand or the technology’s developing capabilities. Research must take an iterative approach to deliver immediately actionable insights so that adoption does not outpace understanding.

Second, we must mitigate AI’s potential to exacerbate biases, privacy violations, access divides, and health inequities. The use of new technologies can lead to unintended consequences. The accuracy and reliability of today’s AI systems depend heavily on the data fed into them, so we must redouble our efforts to ensure these data are as clean and bias-free as possible. 

Third, we must preserve the sanctity of the clinician-patient relationship. AI should enhance the work of physicians, nurses, and allied health professionals—not replace or burden them. It must serve patients and support healthcare professionals in delivering high-quality care. We must ensure that AI technologies are well-considered, thoughtfully deployed, and integrated into clinical workflows to promote interdisciplinary collaboration among the care team and patients.

Finally, we must ensure that AI-enabled health technologies are safe for patients. This brings us back to our recent funding opportunity, which combines two core strengths of AHRQ: its ability to research groundbreaking technologies and its focus on patient safety. It has a twin focus: to understand whether and how breakthrough uses of AI can affect patient safety and to examine the safe implementation and use of AI systems.  

Today, we are on the brink of a new era in which AI will transform healthcare. It’s up to patients, healthcare organizations, and health services researchers to work together to see that it does so in a way that makes care safer, more personalized, efficient, and accessible. With continued dedication to innovation and a commitment to equity and safety, we look forward to the journey ahead.

* In this blog, ‘artificial intelligence’ refers to 21st century AI applications, including generative and predictive models.

Dr. Valdez is director of AHRQ. Dr. Dymek is director of AHRQ’s Digital Healthcare Research Division, where Mr. Chaney is senior advisor for dissemination and innovation.

 Get more ´óÏóÊÓƵnews, sign up for ´óÏóÊÓƵnews  and follow ´óÏóÊÓƵon , , , and .