AI Won’t Replace Medical Monitors, But It Will Redefine How We Work.

Every few weeks, someone asks me a version of the same question: “Do you think AI will replace medical monitors?” It’s a reasonable question, considering the rapid advancements in artificial intelligence. 

These days, AI models are writing reports, summarizing documents, reviewing images, generating code, and performing tasks that once required hours of human effort. Healthcare and clinical research are no exception to this change. 

However, after spending years in oncology drug development, and now using AI as part of my own workflow, I think we’re asking the wrong question. The real question isn’t whether AI will replace medical monitors, but rather how AI will change what it means to be a great medical monitor. These are two very different conversations, and I firmly believe the answer is going to reshape our profession over the next decade.

Medical Monitoring Has Never Been About Processing Information

If we break down medical monitoring to its individual tasks, AI appears incredibly capable. Tasks like reviewing documents, summarizing safety data, drafting narratives, comparing protocol versions, identifying trends, searching medical literature, and generating meeting summaries are all areas where AI is already demonstrating significant value. But that isn’t medical monitoring itself. Those are tasks that support medical monitoring.

The actual job of medical monitoring involves clinical judgment. It’s deciding whether a patient who technically meets eligibility criteria should truly be enrolled. It’s recognizing when several seemingly unrelated adverse events might represent an emerging safety signal. It’s balancing patient safety, protocol integrity, operational realities, and scientific objectives, often with incomplete information. This isn’t pattern matching, but rather it’s decision-making, and decision-making carries responsibility.

AI Will Remove Friction Before It Replaces Expertise

One thing I’ve noticed using AI is that it excels at reducing friction in our daily tasks. These are the small, seemingly insignificant activities that quietly consume time throughout the day, such as reviewing lengthy documents, creating first drafts, finding relevant sections in protocols, summarizing meeting notes, comparing revisions, and searching across vast amounts of information. Anyone who has worked as a medical monitor knows how much time these activities can require.

If AI can reduce these administrative burdens, something important happens. Physicians spend more time thinking, more time discussing difficult clinical questions, more time engaging with investigators, and more time identifying meaningful trends instead of simply searching for them. This shift isn’t replacing expertise, instead it’s creating more space for it.

The Value of a Medical Monitor Is Shifting

For years, part of our value came from being the person who knew where to find the answer. Today, AI can often retrieve information in seconds, changing our expectations. The key differentiator has shifted from access to information to interpretation, understanding context, knowing which data matter and which don’t, recognizing when two similar cases actually require different recommendations, and applying clinical judgment where the protocol is unclear.

These skills are increasingly becoming the hallmarks of exceptional medical monitors. While knowledge remains important, wisdom has become even more valuable.

Pattern Recognition Is About to Change

One of the areas I’m watching most closely is signal detection. Medical monitors spend an enormous amount of time looking for patterns. Are investigators asking the same eligibility question repeatedly? Are certain adverse events appearing more frequently than expected? Is one site interpreting the protocol differently than others? Has enrollment behavior changed? 

These observations often emerge gradually. Sometimes, they’re subtle enough that no single individual notices them immediately. AI has the potential to identify those patterns much earlier. This isn’t because it understands medicine better than physicians, but because it can continuously analyze volumes of information that would be impossible for one person to review manually.

Imagine identifying recurring protocol confusion after the third occurrence instead of the thirtieth. That’s where AI truly becomes interesting. It’s not a replacement for clinical judgment, but rather an amplifier of clinical awareness.

Clinical Judgment Can’t Be Outsourced

This is where I think the conversation becomes more nuanced. People sometimes assume that if AI becomes accurate enough, physicians simply become reviewers of machine-generated recommendations. However, I don’t see it that way.

Every recommendation in medical monitoring has consequences. Patients may or may not receive investigational therapy, doses may be interrupted, safety events may be escalated, and protocols may be interpreted differently. Investigators rely on these recommendations to make real clinical decisions.

Ultimately, someone has to take responsibility, and that responsibility belongs to physicians, not algorithms. AI can inform judgment, but it can’t own it. I believe that’s an important distinction our industry needs to preserve.

Future Medical Monitors Will Need Different Skills

When I think about the next generation entering medical monitoring, I don’t think they’ll simply need to learn protocols, safety reporting, and oncology. They’ll also need to understand how to work effectively alongside AI. This includes knowing when to trust an AI summary, knowing when to verify it, recognizing hallucinations, identifying missing context, understanding the limitations of large language models, and evaluating outputs critically.

These aren’t traditional medical competencies, but I suspect they’ll become increasingly important professional skills. Much like electronic data capture (EDC) systems or electronic medical records (EMR) once became standard, AI literacy will likely become part of everyday clinical development.

Sponsors Should Think Bigger Than Automation

Many conversations around AI focus on efficiency. We’re focused on reducing timelines, decreasing costs, and automating documentation. These are worthwhile goals, but I think the bigger opportunity is improving decision quality.

Imagine AI helping identify patients whose eligibility deserves additional review, highlighting subtle safety trends across global studies, detecting inconsistencies in protocol interpretation, supporting risk-based medical review, or even providing historical context from similar development programs.

These applications don’t simply make work faster. They help teams make better decisions, and that’s where I believe the greatest long-term value lies.

The Human Side of Medical Monitoring Becomes More Important, Not Less

Ironically, the more capable AI becomes, the more I appreciate the parts of medical monitoring that technology can’t replicate. These include building trust with investigators, helping a site navigate a difficult clinical situation, having an honest conversation with a sponsor about an emerging concern, leading multidisciplinary discussions when opinions differ, balancing science with empathy, and explaining uncertainty without creating confusion.

These aren’t technical skills; they’re human ones, and they’re central to what makes medical monitoring effective. While technology can support these interactions, it cannot replace them.

A Final Thought

Every significant technological advancement in medicine has changed the way physicians work. Electronic medical records, advanced imaging, genomic sequencing, and artificial intelligence are just a few examples. But I don’t believe AI diminishes the importance of medical monitors; instead it raises the bar.

As routine tasks become more automated, our value will be defined less by how much information we can process and more by how well we think, communicate, and lead through uncertainty. 

This is an exciting shift, as medical monitoring has never been about producing the longest report or reviewing the most documents. It’s always been about helping sponsors make better decisions, supporting investigators when judgment matters most, protecting patients, and guiding oncology programs through complexity.

AI won’t change this mission; instead it will simply change how we accomplish it. The medical monitors who embrace this future, while holding firmly to the principles of sound clinical judgment, will help define the next chapter of our profession.