When we think about the future of clinical research, it’s easy to picture gleaming technology, seamless data flows, and a world where trials are faster, smarter, and more inclusive. But let’s be honest. We’re not there yet.
That’s exactly the conversation I had in our recent The Future Prescription: Inside the Trends vodcast episode with Guillaume Carbonneau, Vice President of Operational Data Insights at Novo Nordisk. Guillaume is someone I’ve had the privilege of working with for years, and he has a unique way of cutting through the noise and focusing on what really matters when it comes to embedding real-world data (RWD) and artificial intelligence (AI) into clinical development.
This session was packed with insights; from what trials could look like in 2050 to the practical steps teams can take today to overcome inertia. I want to share some highlights here, but I’ll say upfront, you’ll want to listen to the full discussion to really appreciate the depth of Guillaume’s perspective.
Setting the Stage: Ambition vs. Inertia
I opened the conversation with a reality check. We have the tools. We have the data. We even have the ambition. But adoption of data-driven clinical trials is still lagging.
Why? It isn’t the technology. It’s trust. It’s risk. It’s readiness.
Guillaume agreed. In his role at Novo Nordisk, he’s seen firsthand how the real challenge is less about building powerful AI tools and more about whether organizations are ready to use them. He described his team’s mission as leading a cultural transformation, helping colleagues not just access new technologies but actually change the way they make decisions.
That distinction between tools existing and tools being adopted became a theme throughout our discussion.
Precision by the Numbers
One of the first big questions I posed to Guillaume was, what does precision mean in the context of clinical trials in 2050?
For him, precision starts at the planning stage. It means using data to enrich, not replace, human judgment. It’s about combining empirical organizational knowledge with insights from RWD, so decisions are scalable, repeatable, and most importantly measurable.
But Guillaume made it clear that this isn’t just about volume or variety of data. The real challenge is in contextualization. Having massive datasets isn’t useful if they’re poorly structured or dropped into the laps of teams without clarity. As he put it, exposing more data to the wrong audience at the wrong time “will confuse and create mistrust.”
That resonated with me. At TriNetX, we often talk about fit-for-purpose data. It’s not just the size of the dataset but whether it’s curated, contextualized, and trusted. Without trust, the utility erodes.
AI Meets RWD
From there, we shifted into the promise (and frustration) of AI in clinical trials. Too often, projects get stuck in what I call “pilot purgatory”: proof-of-concept initiatives that never scale.
Guillaume shared concrete examples where AI is already making an operational impact at Novo Nordisk. One of the most compelling was what he jokingly called “doing the dirty laundry” or using AI to resolve entities across fragmented datasets. For instance, identifying that Dr. X showing up in 20 different sources is actually the same investigator, and then creating a unified profile that clinical teams can act on. That kind of puzzle-solving may not sound glamorous, but it’s foundational for better site and investigator selection.
He also described experiments with generative AI to design future clinical trials. By analyzing historical protocols, including not just their structure but also their operational outcomes, his team is exploring ways to create smarter, more efficient study designs. Imagine an engine that learns from thousands of past studies to suggest new endpoints or inclusion criteria. It’s not replacing scientists, but it is expanding the possibilities for innovation.
From Interesting to Indispensable
What does it take for AI to move from being an interesting tool to an indispensable one? Guillaume’s answer was striking. He drew from his prior experience at Walt Disney World, where customer understanding was everything, and translated that concept to clinical research.
The true value of AI, he said, is in uncovering the 95% of opportunities we don’t currently see such as research-naïve physicians, new institutions, or patient populations outside of the well-trodden trial networks. Identifying those hidden opportunities could be game-changing for the industry.
That’s when AI stops being a novelty and starts becoming central to growth and competitiveness.
Breaking Through Inertia
Of course, none of this matters if organizations remain stuck in their traditional ways of working. Guillaume and I talked about how risk aversion often leads to analysis paralysis—overvalidating, overtesting, and ultimately hesitating.
His approach? Don’t shy away from “boiling the ocean.” With the computing power we now have, it’s possible to analyze millions of potential physicians or sites quickly. Even if some avenues don’t materialize, the exercise builds curiosity and confidence among teams. And that curiosity is often the spark needed to overcome inertia.
I couldn’t agree more. In my own experience, I’ve seen how giving teams rapid, tangible results builds momentum. When the data speaks clearly and quickly, the fear of moving forward begins to fade.
ROI and Trust
Another important theme was return on investment (ROI). In clinical research, ROI isn’t just about cost savings or speed. It’s also about outcomes, risk reduction, and patient experience.
Guillaume shared an example that really stood out: using AI and modeling to challenge the myth that diversity slows down recruitment. By targeting physicians who naturally treat diverse populations, his team was able to improve representation in trials without slowing enrollment. In fact, in some cases, recruitment became faster. That’s the kind of ROI that can’t be ignored: better science, faster timelines, and more representative outcomes.
But alongside ROI is the issue of trust. Guillaume was candid that his teams are still in the process of building trust in AI-generated recommendations. Today, it’s about demonstrating that the engine can understand historical data. Tomorrow, it will be about trusting it to recommend new study designs. That shift, from verification to recommendation, will be one of the most important leadership challenges of our time.
Leadership’s Role
That brought us to the role of leadership. For me, leadership is about creating the time and space for innovation, building patience, and fostering trust. Guillaume expanded on that, emphasizing the need for leaders to inspire teams to embrace a different kind of decision-making.
Traditionally, decisions have been made based on personal networks and empirical knowledge. We trusted people because we knew them. Now, trust increasingly has to extend to data and to AI-generated insights, things we may not fully understand. Leaders need the courage to make that shift possible.
Looking Ahead to 2050
So, what does 2050 look like if we get this right?
Guillaume’s vision was bold: shrinking processes that currently take months into workflows that happen in weeks. Imagine AI engines that can take a protocol, generate optimal site and investigator plans, and even draft standard agreements, all while maintaining scientific rigor and compliance.
The goal isn’t to replace people but to eliminate unnecessary administrative drag so patients can access treatments faster. And that’s the kind of future worth striving for.
Closing Thoughts
As we wrapped up, I asked Guillaume two rapid-fire questions. His answers?
- One myth about AI in clinical research that needs to die: Humans always know better than AI.
- One action organizations should take now: Be curious. Don’t cling to the good old days. Embrace the change because it isn’t going backward.
Those two ideas, humility and curiosity, might just be the keys to unlocking the future of clinical trials.
Access all of the insights. Watch the full vodcast episode with Guillaume Carbonneau on data science and AI in clinical trials and explore The Future Prescription report for an in-depth look at the five trends shaping healthcare and clinical research.
About Steve Kundrot
Steve is a technology and business leader with over 20 years of experience in clinical research, health analytics, consulting, and software development. As Chief Operating Officer, he oversees TriNetX’s core operational functions and leads the development of a unified product roadmap designed to revolutionize clinical research and accelerate drug development by optimizing clinical trial design, enhancing post-market safety, and delivering research-grade data and evidence that enable and expedite regulatory approvals.
About Guillaume Carbonneau
Guillaume Carbonneau is an accomplished leader in operational insights, specializing in clinical development business problem-solving and decision-making. With two decades in the pharmaceutical industry, he has spent the last 10 years leveraging technology to enhance patient outcomes and foster innovation.
As Vice President of Operational Data Insights at Novo Nordisk, he leads a transformative team focused on data-driven decision-making and developing the StudyHub platform. His team’s expertise optimizes clinical trial design, planning, forecasting and execution, ensuring AI-generated insights have real-world impact. They are fostering a culture where data drives decisions across the portfolio.
Before Novo Nordisk, Guillaume was Head of Health Data Insights & Design at Novartis. There, he utilized cross-disciplinary experience to develop data-driven methodologies, incorporating profiling methodologies and design thinking to enhance product desirability. His collaborative approach led to innovative healthcare solutions in partnership with external stakeholders, underlining his commitment to transforming clinical practices.
Guillaume’s passion for technology, data and AI, continues to drive innovation in data insights, creating actionable outcomes that deliver tangible results.
Guillaume is not an official spokesperson for Novo Nordisk, and his comments reflect his personal perspective and experience, informed by his role at the company.