Are Sites Even Using AI?: How Sponsors And CROs Can Support Uptake

By Maria P. Ladd, site operations consultant

AI is no longer just a far-off concept in clinical research — it’s here, and it’s growing. Whether you love it, fear it, or use it without much thought, AI is embedded in our daily work and culture. Everyone from tech enthusiasts to skeptics is asking: What should we be doing with it? And how? Some of that may be FOMO. Because, hey, no one wants to miss out on what’s clearly the next big thing.

But who benefits the most from AI in clinical research? Is it primarily sponsors? What about CROs? Does it even involve clinical research sites yet? And what can AI do to make the day-to-day operations of a site more efficient, accurate, and less burdensome?

For sites, the question isn’t just what AI can do — it’s whether it’s even relevant to them yet. Right now, the answer is complicated.

Sites Are Still On The Sidelines

Ask the average site’s staff about AI, and you’re likely to get a blank stare — or silence. In conference sessions, many hesitate to ask questions, fearing they’ll sound out of touch or behind the curve. The industry talks a lot about innovation, but there’s often little guidance on how to apply it where the rubber meets the road: the clinic, the site, the staff.

Most sites don’t know what AI tools exist for them, who’s supposed to pay for them, or how they fit into already-stretched workflows. Site directors and coordinators aren’t resistant to change, but they are rightly skeptical of anything that adds complexity without clearly saving time or improving outcomes.

What AI Can Do At The Site Level (Right Now)

So, what can AI do for a clinical research site? The standout use case is patient recruitment. AI tools can scan EMRs or site databases to flag potential participants based on inclusion/exclusion criteria, saving hours of manual review. Some platforms even use machine learning to rank candidates by likelihood of participation.

Let’s say a site is running a trial for moderate-to-severe asthma. Instead of manually sifting through charts or relying solely on physician referrals, a coordinator could use an AI tool to flag patients with qualifying protocol criteria. The coordinator gets a short list to call, instead of a haystack to dig through. It means real-world time saved in continuing resource constraints. But beyond that? The list of useful site-specific AI tools gets thinner.

While sponsors and CROs are exploring AI for protocol design, risk-based monitoring, and predictive analytics, these innovations rarely make their way to the research site in a meaningful way. Many sites are still tasked with juggling multiple sponsor portals, logging into different EDC systems, and transcribing data by hand from paper source documents.

What A Quick Search Won’t Tell You

A quick online search for “AI in clinical research” brings up terms like natural language processing, digital biomarkers, and trial simulation modeling. But nearly all are sponsor- or CRO-facing, and these innovations are designed to help upstream decision makers run trials more efficiently or bring products to market faster.

The site — the place where human interaction, data collection, and patient care happen — is historically an afterthought in tech planning. It’s the last mile in the clinical research pipeline, but the most overlooked when it comes to technology investment.

What’s Missing: Practical, Funded Site-Facing Tools

Outside of recruitment, there’s little practical incentive for sites to invest in AI-powered tools. Many of them simply don’t have the budget or IT infrastructure. Sponsors and CROs, who often dictate the tools and platforms used in a study, hold the keys to what gets adopted — and what doesn’t.

But there’s untapped potential, and it could soon be realized just by starting a conversation.

What if sponsors partnered with site representative organizations to co-develop AI tools tailored for use at the site level? What if AI were built into protocol design with site workflows in mind (and site input), not just data points? Picture a platform that auto-generates visit windows based on the site and staff calendar or flags protocol deviations in real time.

There’s room for an even broader vision: AI-powered feasibility assessments that include feedback loops from sites or risk-based budgeting that accounts for the cost of technology adoption. These are the kinds of innovations that could improve site performance and trial quality across the board. It’s well past time for sponsors and CROs to play a bigger role in making the process as efficient as possible for clinical research sites; there’s a lot at stake.

What Sites Bring To The Table

What sites contribute to the research ecosystem, and cannot be imitated, is the human connection. Sites build trust with participants; they provide support after procedures; they are the human layer that can’t be replaced by even the most sophisticated AI.

Beyond the protocol, monitoring visits, and documents in a binder, site staff are at the people level of research:

• A coordinator knows that Mrs. D might enroll because her husband had the same condition.
• A site calls Mr. G the day after an infusion to check in — not because it’s in the protocol, but because it’s the right thing to do.
• A study nurse holds a participant’s hand during a difficult visit.

These are not AI problems; they’re human experiences, and they are the foundation of trustworthy, high-quality research.

AI Can Help But Not Replace

As AI continues to evolve, its role at the site level will grow. We’ll likely see more intelligent scheduling, error detection, and source-to-EDC transcription. But those tools must be built for sites, not just bolted on.

AI should amplify the value of the site — not attempt to automate it away. It should reduce the administrative burdens so coordinators can spend more time with patients. It should catch compliance issues early so staff can focus on care — and it should be widely implemented with funding, training, partnership, and support.

Final Thought: Humanity

Clinical research sites are the beating heart of every trial. They are the ones doing the hard work, the emotional labor, and the clinical care that make data collection possible. The people at these sites are showing up every day — smiling, coordinating, problem-solving, and advocating for patients.

Yes, AI is powerful. Yes, it’s the future. But without empowering the research site, we’re building a high-tech vehicle without providing fuel.

It’s time for sponsors, CROs, and tech developers to look downstream, not just upstream, and invest in the people who make research possible. AI may shape the future of trials, but it will never replace the human connection that drives them forward.

About The Author:

Maria P. Ladd has been in the clinical research industry for close to 20 years. With her beginnings rooted in an administrative role in start-up, eventually leading to managing a large global team of study support staff in the CRO world, Maria is a passionate research site advocate, site operations consultant, and co-founder of the Clinical Research Site Collective. Maria can be found on LinkedIn where she lends a strong voice to the clinical research community at large.