NOTE: This article was originally published in September 2026 as part of our PharPoint of View LinkedIn newsletter. Each month, our PharPOV series shares perspectives from across the PharPoint team: operational insights, trending topics, and stories about the behind-the-scenes collaboration that keeps studies moving. This article shares the perspective of Paul Johnson, PharPoint’s Executive Director of Strategy Development.

AI in Clinical Research: A Powerful Tool, But Not a Substitute for Judgement

Earlier this year, my colleague, Jayme Swinson, presented a compelling PharPOV on the impact of AI on clinical research. I highly recommend reading his piece as it clearly lays out the benefits and risks of implementing AI technologies as part of clinical trial workstreams. Today, I’d like to talk about the real-world impact AI may have on upcoming clinical trials from a strategic perspective.

A few months ago, I attended a symposium hosted by one of our partners highlighting the promise and capabilities of a new AI-enabled data analysis tool.  The presentation featured appealing graphics and memorable soundbites, all promising an experience that could serve as a force multiplier and accelerator, trimming weeks, or even months, from study startup, database development, and enrollment.  It looked and sounded impressive, but it seemed to be missing one important factor: the human element.

In the days following the meeting, I found myself thinking about the new environment we are entering. AI is rapidly reshaping how we think about clinical research.  Across the industry, we are seeing meaningful progress, from faster data processing to more efficient trial operations.

There is no question that AI has the potential to improve how we develop therapies and bring them to patients. I view that as a positive development. But I also believe we need to approach this moment with discipline.

As we integrate AI more deeply into clinical development, we have to be equally clear about where it adds value and where its limitations require us to pause, question, and apply human judgment. Because, in the end, clinical research is not just a technical exercise but a human responsibility.

Where AI Is Clearly Adding Value

In my experience, AI is already delivering real benefits, particularly in areas that are data-heavy and operationally complex. It is helping teams analyze historical trial data to improve site selection and feasibility assessments.  It is streamlining data management processes, identifying anomalies more quickly, and accelerating timelines that were once constrained by manual workflows.  Administrative tasks that previously consumed valuable time can now be automated or significantly reduced.

Speed and efficiency are critical in drug development – but the real opportunity is not just about doing things faster; it is about doing them better. The true value of AI lies not in replacing expertise, but in enabling better, more informed decisions.

At its best, AI helps reduce noise.  It allows experienced teams to focus their attention where it matters most: interpreting results, managing risk, and making decisions that require experience and judgment. That is where the real value is created.

Where We Need to Be Careful

For all of its strengths, AI has real limitations, and in clinical research, those limitations are significant.

AI models are only as good as the data behind them.  If the underlying data is incomplete, biased, or not representative, the outputs will reflect those same issues.  That is not a theoretical concern; it is a practical one in an industry where patient populations and standards of care continue to evolve.

More importantly, AI does not understand context in the way experienced professionals do. It can identify patterns and correlations, but it cannot fully interpret what those findings mean in the context of patient safety or clinical outcomes. AI can accelerate clinical research, but it cannot replace the judgment required to safeguard patient outcomes.

When teams are evaluating adverse events, adjusting a protocol, or making decisions that directly affect patients, there is simply no substitute for clinical experience.

That is not a limitation we can overlook.

The Risk of Getting This Wrong

One of the risks I see emerging is not the failure of AI, but the possibility of over-reliance on it.

As these tools become more sophisticated, there is a natural tendency to trust the outputs, especially when they are presented with speed and confidence. But speed should never be confused with certainty.

Efficiency gained through AI is only meaningful if it is matched by confidence in the quality and integrity of the results. If we prioritize efficiency without fully validating what we are seeing, we risk introducing new vulnerabilities into the process, including missed safety signals, flawed assumptions, or decisions that cannot be adequately explained or defended.

In a regulated environment, that matters. In a patient-centered environment, it matters even more.

Accountability Still Rests with Us

At the end of the day, clinical development is built on accountability. Every decision has an owner, every outcome has implications, and every trial must stand up to regulatory scrutiny and ethical responsibility.

In clinical development, accountability cannot be automated; human oversight remains essential. AI can support our work, but it cannot take responsibility for it. It cannot sign off on a decision, stand behind a result, or answer for an outcome.

That responsibility remains with the people leading the work.

And it should.

Finding the Right Balance

I do not believe this is a question of whether we should use AI.  We should, and we will continue to do so. The better question is how we use it responsibly.

For me, that comes down to maintaining a clear “human-in-the-loop” approach.  Use AI where it is strongest: processing data, identifying trends, and improving efficiency.  But ensure that experienced professionals remain firmly in control of interpretation, decision-making, and oversight. That balance is what allows us to move forward with confidence.

The future of clinical trials will be shaped by AI but defined by the people who guide its use. Technology will continue to evolve – that is inevitable – but the integrity of clinical research will always depend on how we apply it and the standards we choose to uphold.

The PharPoint Perspective

From our standpoint, this balance is especially important.

We sit at the intersection of sponsors, sites, and patients.  That position comes with both opportunity and responsibility.  We have the ability to bring innovation into the clinical process, and we also have an obligation to ensure it is applied in a way that strengthens quality and integrity rather than compromising them.

At PharPoint, we see AI as an important tool, but not a defining one. Our focus remains on experienced teams, strong oversight, and disciplined execution, supported by technology where it adds value.

AI will continue to play an increasingly important role in clinical research.  It will help us work more efficiently, process more data, and potentially bring therapies to patients faster. But we should be clear about one thing: clinical development is not just about information; it is about judgment, accountability, and trust.

If we get that balance right, AI will make us better. If we get it wrong, it will simply make us faster at making mistakes. That is why, in my view, the path forward is not about maximizing automation, but applying innovation with discipline.

As the role of AI in clinical research continues to expand, the conversation is just beginning.  I welcome the opportunity to connect with sponsors, partners, and industry colleagues to share perspectives on how we can harness these tools responsibly while maintaining the high standards our industry demands. If you are exploring how to navigate clinical development in this new environment without compromising quality or oversight, let’s start a conversation.


About the Author

Paul Johnson, Executive Director of Strategy Development

Paul Johnson has over 35 years of healthcare experience, including over 25 years in clinical research with a foundation within clinical operations.

Paul joined PharPoint’s clinical operations team in 2018. In his current role, Paul leads strategy and business development and plays a central role in early client engagements and vendor management. His therapeutic experience includes oncology, CNS, respiratory, and early-phase research.

Paul has served as a conference speaker on the topic of clinical trial budget management and was a contributing author to a whitepaper focused within study decentralization. He holds a BS in Biology and Natural Science from Midland University.


RELATED RESOURCES

It Ain’t Much, But It’s Honest Work: IT Edition | PharPOV Series

Exec. Director of IT Sean McRae talks about the honest work involved in data protection.

Fit Matters – CROs are Not All the Same | PharPOV Series

When it comes to CRO selection, fit matters, and not every vendor is the same. Hear from PharPoint co-founder Audra McRae.

In Clinical Research, AI Cannot Be A Substitute for Judgement | PharPOV Series

Exec. Director of Strategy Development Paul Johnson talks about the use of AI in clinical research in this month's PharPOV.

Assessing Quality Culture | PharPOV Series

How mature is your organization’s quality culture? Hear from PharPoint's Director of Quality Assurance.

Navigating Complexity in Pulmonary Trials | PharPOV Series

Exec. Dir. Strategy Development Paul Johnson kicks off the first article in a series on pulmonary trial design.

AI in Clinical Research | PharPOV Series

How can we make sure artificial intelligence (AI) is used ethically and responsibly in clinical research? Hear from PharPoint's VP of Operations.
Written by: Theresa Hegar