Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Promises and pitfalls of artificial intelligence in healthcare

The evolution of AI can mean exciting things for healthcare, but like any new technology, it comes with red flags.

[Images: Adobe Stock / Iryna]

Profile photo of Devin Jopp

CEO · Association for Professionals in Infection Control & Epidemiology (APIC)

Going back in time to 1995-2000, the internet was indeed the next “big thing” as we watched it move from obscurity to becoming engrained into our society. Back in 1999, I served as the Chief Information Officer for the Health Insurance Association of America (HIAA), and while the dot-com era was well underway, we still had no idea exactly how large an impact the internet would have. And here we are, nearly 25 years later, with the internet having eclipsed our wildest imaginations.

This brings us to artificial intelligence (AI), which is defined as “…the simulation of human intelligence by a system or a machine… [with the goal of AI to] … develop a machine that can think like humans and mimic human behaviors, including perceiving, reasoning, learning, planning, predicting, and so on.” Similar to the trajectory of the internet in the ’90s, the question is not whether AI will be important or not, but how important and when.  

An article published by Time in November 2023 makes the case that AI is not slowing down and, in fact, is accelerating immensely in its ability to outperform humans on a number of tasks, from reading to language comprehension. Even when considering empathy, which we would consider a uniquely human trait, AI is making gains.  

There are myriad possible AI applications in healthcare, including improving patient safety and quality, accelerating scientific discovery, improving physician and nurse decision-making, improving the patient experience, and improving healthcare administration.  According to HITRUST, numerous applications of AI are already being deployed in various aspects of healthcare, including:

  • Drug discovery and development
  • Predictive analytics and risk assessments
  • Diagnosis and disease detection
  • Treatment planning and personalization
  • Administrative efficiency
  • Telemedicine and remote monitoring

Today, we are seeing a very early version of AI focusing on large language models. But as the technology becomes more sophisticated, so will the ability to incorporate generative models that can produce non-text responses (e.g., images and videos). We’re just touching the surface today.

Healthcare delivery systems will most likely focus their initial efforts on deploying AI to leverage information already being captured and linking multiple sets of data to create new knowledge and improve decision-making. 

Today, the U.S. healthcare system is drowning in vast amounts of data. In hospitals alone, this can be data from electronic health record systems, administrative billing data, patient portals, remote monitoring systems, e-prescribing systems, etc. The veracity of data and sources is increasing, with wearables, large regional qualified healthcare information networks (QHINs), and eventually, genomic data.   

As AI continues to become more advanced, we will see the use of AI increase in diagnosis and disease detection, from both structured data sources (e.g., EHRs) to unstructured data sources like radiology, blood tests, and medical devices.  

Further into the future, with the increasing sophistication of AI and tools like imaging and robotics, there are even greater applications that may be adopted by healthcare. Imagine cameras in facilities (or even in patients’ homes) detecting the potential for falls or alerting staff when a patient at risk for bed sores hasn’t been moved. Eventually, you could even imagine utilizing automated robots to clean spaces or using imaging and AI to ensure proper cleaning protocols are followed. 

But this optimism for AI in healthcare also comes with roadblocks and caution flags. For example: 

  • Accuracy And Bias: AI findings may be inaccurate at times, yielding false results, or biases may be introduced based on input variables.
  • Switching Workflows: The challenge of upheaving existing workflows, deploying new technology, and training poses significant headwinds to AI.
  • Legal Issues: Relying on AI-generated knowledge that may not be accurate could lead to legal issues.
  • Patient And Healthcare Worker Privacy: Ensure healthcare workers and patients maintain their privacy when many of these technologies capture an individual’s identity (e.g., facial recognition imaging data).
  • ROI: It’s unknown whether new AI tools will actually improve outcomes, drive revenue, or improve efficiency.
  • Trust And Fear: Implementing new tools like AI brings with it the challenge of building trust in the new tool and addressing what fears may be present.
  • Equity: Implementing these new AI-driven tools will cost financial resources, which inherently creates the have and have-nots in the healthcare ecosystem.
  • Complexity, Difficulty, And Unpredictability: Implementing AI-driven tools is limited by human trial and error, which can lead to challenges in replicating success. Models must continually be re-trained with new information.

So, what are some tips for healthcare leaders considering the future of AI for their organizations?

  • Start Educating Your Organizations On AI: Consider bringing in peers who have success stories, or others who are familiar with AI advances, to stimulate thinking and help level set an understanding among staff and board members.
  • Prioritize: Like any good initiative, prioritize AI activities.
  • Consider Risks And Roadblocks: Evaluate what risks and roadblocks exist around AI strategies, and how the organization can best address them. Don’t forget to evaluate regulations on AI.
  • Start Small: Look for ways to implement pilot programs and evaluate efficacy.
  • Incorporate AI Into Strategic And Operational Plans: Review strategic and operational plans to determine if and how AI will be positioned in the organization.
  • Stay Nimble And Open: What we know about AI will change, and it is important to be open to adapting and changing with its evolution over time.

Implementation of AI in healthcare will be a marathon, not a sprint. To successfully harness the power of AI, healthcare leaders must evaluate real-world use cases and demonstrate how it can help the organization meet the triple aim of improving the experience of care, improving the health of populations, and reducing per capita costs of healthcare. 

While it’s hard to see exactly what the crystal ball holds for AI, the adoption of the internet probably holds a few clues on what’s likely ahead.


Dr. Devin Jopp is Chief Executive Officer of the Association for Professionals in Infection Control & Epidemiology (APIC).