Generative artificial intelligence (GAI) technologies have grown at an astounding pace. Less than 3 years after the launch of ChatGPT, the chatbot already receives more than 5 billion visits each month.
In addition, the race is on to integrate GAI tools into all kinds of professional settings — including the clinical lab. For example, some experts believe GAI has the potential to help laboratories streamline and facilitate operations, management, and research. But at this stage, it can be difficult to differentiate between real possibilities and public relations spin.
At a roundtable discussion led by Li Zha, PhD, DABCC, NRCC, associate director of clinical chemistry at Boston Children's Hospital, attendees at ADLM 2025 (formerly the AACC Annual Scientific Meeting & Clinical Lab Expo) will learn how to separate the wheat from the chaff when it comes to GAI.
“The overall goal of the session is … to bring a really up-to-date, high-level overview,” said Zha. He will start by reviewing a series of high-quality articles assessing the use of AI for diagnosis and image analysis. Then Zha will facilitate an interactive discussion about a series of scenarios highlighting the capabilities of currently available GAI technologies.
The group will discuss where GAI is already being used in healthcare — like hospital chatbots, for example. “That can essentially serve as an entryway to this,” he said. Electronic health record companies are also integrating GAI to streamline workflows, assess reporting processes, and summarize notes.
“[Use cases] will only increase in terms of frequency,” Zha said. “From the laboratory standing, arguably I really believe down the road it’s good for laboratory experts to be helping in that endeavor and validation.”
He’ll also talk about potential drawbacks and risks, especially when handling sensitive patient information. The last thing a clinical laboratorian wants to do is to create a HIPAA violation by drawing from nonanonymized patient data.
In true roundtable fashion, Zha plans to end with a practical discussion of where GAI might fit into clinical laboratories. “As a pivot point for interaction, the goal is to really see where … as practitioners in laboratory medicine, what do we need versus what do we have,” he said, and “what major limitations do we need to bridge to really make these tools deliver these promises?”
He hopes attendees will leave the session with a stronger grasp on where GAI is as a technology right now and “take away some ideas that they can experiment with in their labs. I want the field to look into the future and think about what we can do … to really make the tools work for us as opposed to it being thrown at instrumentation and tools.”
Instead of taking a back seat to how GAI is applied to clinical laboratories, clinical laboratorians should be in the driver’s seat, Zha said. “We need to be thinking about the scenarios where we can most benefit from using these tools,” he said. “We are the producer of the laboratory data, and we know a lot about how different pieces are and aren’t compatible.”
Despite all the buzz around it, GAI is still a young technology, Zha added. “It’s like the early days of Wikipedia … where students in middle school were told not to really trust everything on it.” But it has become a better, more dependable resource with time, as GAI could be.
Jen A. Miller is a freelance journalist who lives in Audubon, New Jersey. +Bluesky: @byjenamiller.bsky.social