CROs, AI, And The New Economics Of Outsourcing

By Dan Schell, Chief Editor, Clinical Leader

For the past few years, there has been a quiet but persistent question hanging over the clinical trial outsourcing model. It’s not about demand. It’s not about funding. It’s about whether the work itself might start to require fewer people.

That question came into focus during a recent LinkedIn Live discussion between Castor CEO Derk Arts and CRO consultant Joel White, combined with White’s latest earnings recap. Taken together, they highlight a tension the industry hasn’t fully resolved yet. The outsourcing model was built on labor. The next phase of clinical trials may be built on efficiency. Right now, though, you wouldn’t know anything is changing.

A Strong Market — And A Subtle Shift Beneath It

White’s Q4 recap shows an industry that has rebounded in a meaningful way. After a difficult stretch in 2024 and early 2025, biotech funding has come back, RFP activity is up, and CRO bookings are strong. Decision timelines are tightening, and more work is flowing through the system again. That matters because, historically, more funding leads directly to more trials, and more trials lead to more outsourced work. The connection between those variables has been one of the most reliable dynamics in clinical research. When I followed up with White via email after the LinkedIn Live, he explained that he still believes that more funding-more trials-more outsourcing relationship will hold, even as efficiency improves. “I personally believe that reducing the cost to run a trial would lead to more trials being run, negating the doomsday scenarios about the business model.”

That view assumes demand will expand as costs come down. It’s a logical argument, and for now, the data supports it. CRO revenues are growing alongside increasing trial activity, not diverging from it. But there is still this concern about how AI will affect everything.  

Efficiency, Pricing, And Where The Value Flows

The fear is straightforward. If AI, automation, and better workflows reduce the amount of work required per trial, then over time the outsourcing model should feel that impact. Less work should eventually translate into less revenue. White doesn’t dismiss that idea, but he sees it playing out gradually and unevenly. “It would likely first pressure revenue growth (modestly) while benefiting profit margins, while over a longer timeframe decrease the cost of drug development per trial,” he said.

That nuance matters because CROs are no longer simply billing for time. Most services today are structured around fixed units or milestone-based pricing. A monitoring visit, a data deliverable, or a study milestone carries a defined price regardless of how efficiently it is executed. If a CRO can complete that work faster or with fewer resources, it doesn’t immediately lose revenue. It improves margin.

That fact made me think of the age-old argument that CROs are incentivized to resist efficiency because it reduces billable work. When I asked White about this, he was blunt: That argument is simply wrong. “Rule 1 as a CRO is to continually win new work,” he told me. “To do so in a cutthroat competitive environment, you have to be price competitive, and to be price competitive while remaining profitable, your operations have to continually become more efficient. I cannot stress enough how ultra-competitive the CRO world is, where the vast majority of awards are made via competitive RFP processes with multiple rounds of tough negotiations. You cannot resist operational efficiency and stay competitive in this environment.”

That said, efficiency gains do not benefit all stakeholders equally, at least not at the same time. White’s view is that value shifts over time: Early adopters — typically CROs and sites — capture the initial gains, but as adoption becomes widespread, competitive pressure pushes more of that value back to sponsors. That dynamic becomes even more interesting when you layer in another structural shift happening in parallel.

Will Large Site Networks Compete With CROs?

The rapid consolidation of independent clinical research sites into large, corporate multisite networks — and the formation of organizing bodies like AMRC — has prompted questions about whether these entities could absorb functions historically owned by CROs, thereby eroding a portion of the CRO outsourcing market.

At first blush, I understood this concept. When organized and large site networks negotiate budgets directly with sponsors, CRO pass-through revenue could shrink. Further, when sponsors engage site networks with standardized SOPs and centralized contracting, the CRO’s traditional site selection and management role could diminish. Of course, the key word in both of those sentences is “could.”

Frankly, I’m just not really buying this theory.

Sure, some large site networks can be great at patient recruitment, retention, and data collection, but they don’t perform regulatory strategy, pharmacovigilance, biostatistics, or global multicountry coordination, and they are unlikely to develop those capabilities at scale. At best, I’d say we’ll see large site networks start to chip away at CRO involvement in site management and pass-through revenue. It will be more like a shift in where CROs play and how they make their money — not a replacement of the FSO model.

What Would Actually Signal A Problem?

For all the discussion around AI, automation, and shifting value chains, the most important question is still a simple one: What would real disruption actually look like in the numbers?

White has a clear answer. “I would need to see CRO revenues fall as clinical trial starts stay steady or increase,” he said. “If trial starts increase 5% but CRO revenues flatten or decline, over multiple quarters, I would get very concerned.”

That is the signal to watch. Not headlines, not product announcements, not conference presentations, but a sustained divergence between trial activity and CRO financial performance. So far, that divergence has not appeared. Trial starts are increasing, and CRO revenues are rising with them. Even as new technologies promise to reduce manual work, those gains have not yet translated into measurable revenue pressure.

Part of the reason is timing. Clinical trials move slowly, and sponsors remain cautious about introducing new technologies into active studies. Many of the tools expected to drive meaningful efficiency — eSource, EHR-to-EDC integration, reduced SDV — are still in early stages of adoption at scale. Their impact, while widely discussed, has not yet fully worked its way into operational or financial results. That lag creates a window where both things can be true at once. The technology is advancing, and the business model remains intact.

Make no mistake; demand for clinical trial services is growing, not shrinking, and the complexity of modern studies continues to require specialized expertise that sponsors are unlikely to internalize at scale. What is changing is how value is created and captured. I’d say the outsourcing model is being … rewritten, albeit gradually, unevenly, and in ways that may only become obvious once the numbers start to tell a different story.