Clinical trial sites have become the operational backbone of increasingly complex studies. Yet while expectations continue to grow, site resources often do not.
Protocol complexity, increasing data requirements, staffing challenges, and administrative workload continue to place pressure on site teams responsible for delivering high-quality studies and positive patient experiences. Sites are expected to manage growing operational demands while maintaining compliance, data quality, and ambitious study timelines.
Against this backdrop, artificial intelligence (AI) is emerging as a potential solution. But the real question is not whether AI can be applied at the site level—it is whether these tools can reduce burden without introducing new complexity.
While much of the industry conversation has focused on future possibilities, sites are increasingly exploring practical applications that may help improve efficiency and reduce repetitive administrative work. As one site leader put it during a recent discussion, any technology introduced into the site environment has to genuinely reduce burden rather than create additional operational friction. That captures both the challenge and the opportunity.
Supporting Sites in an Increasingly Complex Trial Environment
Most sites are not looking to replace people with technology. They are looking for practical ways to help already stretched teams manage growing workloads more efficiently.
Reducing Administrative Burden
Site coordinators and investigators spend considerable time on administrative activities that are essential to trial delivery but can also be repetitive and time-consuming. Tools powered by AI may help support activities such as:
- Summarizing protocols and study documents
- Drafting non-regulatory communications
- Supporting patient pre-screening activities
- Organizing recruitment workflows
- Prioritizing data queries
- Assisting with multilingual communication
- Identifying missing documentation
- Supporting internal training and knowledge sharing
Many of these improvements may seem small individually, but for teams managing multiple studies, patients, and competing priorities, they can add up quickly.
Consider a site coordinator reviewing protocol amendments, responding to routine queries, coordinating patient visits, and maintaining study documentation across several active studies. If technology can reduce the administrative effort required for these activities, it creates more time for patient engagement and study oversight. If it simply introduces another platform to learn and manage, it may add to the burden rather than reduce it.
That distinction is important. The value of any new technology should ultimately be measured by whether it helps sites operate more effectively.
AI should be viewed as a support tool, not a replacement for investigator expertise or coordinator judgment. Clinical research still relies heavily on human oversight, clinical decision-making, and patient trust.
Supporting More Connected and Data-Driven Trial Delivery
Another reason this discussion is becoming increasingly important is the growing focus on more connected and data-driven approaches to clinical trial execution.
Initiatives such as the FDA’s Real-Time Clinical Trial (RTCT) Initiative reflect a broader industry movement toward accessing and using study data more effectively during the conduct of a trial rather than waiting until key milestones or study completion to identify issues and make decisions.
For sites, this creates both opportunities and challenges.
As more information becomes available throughout the lifecycle of a study, the challenge is no longer simply collecting data—it is identifying what matters, prioritizing actions, and responding to issues in a timely manner without creating additional burden for site teams.
This is where AI may have an important role to play.
By helping organize, review, and interpret growing volumes of operational and clinical information, these tools may support more proactive trial management while allowing investigators and site staff to remain focused on patient care and study delivery.
At the same time, the same principle applies: technology should simplify trial delivery, not make it more complicated.
Whether supporting administrative workflows or helping teams manage increasingly data-rich clinical trials, the measure of success will not be how sophisticated the technology is—it will be whether it helps sites, investigators, sponsors, and patients participate in clinical research more effectively.
Adoption Must Be Practical
In a recent discussion with sites, what stood out was that the conversation wasn’t really about the technology itself—it was about trust.
Sites were interested in where these tools could help, particularly in reducing administrative burden and supporting day-to-day operations. However, they were equally focused on understanding what guardrails should exist around their use.
One topic that came up consistently was the use of public AI tools.
Sites drew a clear distinction between using AI to support internal operational activities and entering sensitive information into publicly available platforms. While using AI to organize documentation, support recruitment planning, identify operational risks, or streamline internal processes was generally viewed as appropriate, entering confidential information was not.
There was strong alignment on a critical point: proprietary sponsor data, CRO information, and any patient-related or confidential study data must not be entered into open or public AI platforms under any circumstances.
Such information should only be used within approved business enterprise systems. These are secure, organization-controlled environments designed to protect sensitive data. They typically include:
- Access controls that limit who can view or use information
- Data encryption and security protections
- Audit trails to track how information is used
- Governance aligned with regulatory and compliance requirements
In contrast, publicly available AI tools do not provide the same level of control, transparency, or assurance around how data may be stored, used, or accessed.
Several participants were clear that this represents a non-negotiable boundary. Any use of AI in clinical research must prioritize confidentiality, data integrity, and regulatory compliance.
As adoption evolves, sites are seeking clearer guidance on:
- What information can and cannot be used in AI tools
- Which platforms are approved for use
- What safeguards and oversight are required
The goal is not to discourage adoption, but to ensure that these tools are used in ways that are safe, compliant, and practical within real-world site operations.
AI Adoption Requires Alignment Across the Clinical Trial Ecosystem
While investigative sites are where many of these workflows become operational, adoption cannot happen in isolation.
Sponsors, CROs, principal investigators, sites, and technology providers all play a role in determining how these tools are introduced into clinical research environments and whether they genuinely improve trial delivery.
One thing I have learned from working with sites across a wide range of studies and regions is that what looks good in a presentation does not always work in practice. Sites are ultimately the ones who have to make these processes work every day.
From my perspective, one of the most important roles a CRO can play is helping determine whether a new technology will genuinely help sites or simply give them one more thing to manage.
Technology by itself is rarely the answer. The real challenge is how these tools fit into the realities of running a study.
Principal investigators also remain central to successful adoption. While these tools may support administrative efficiency, investigator oversight, clinical judgment, and accountability remain fundamental to patient safety and trial integrity.
Patient trust is equally important. As these technologies become more visible within clinical research operations, transparency around how they are used will remain critical to maintaining confidence in the research process.
What Sites Need from AI
Based on the conversations I have had, sites are looking for a few things.
They want tools that solve real problems rather than introduce new ones. They want clarity around what information can and cannot be used and what expectations sponsors and CROs will have as adoption evolves.
They want confidence that technology will support investigator oversight rather than complicate it. And they want solutions that fit naturally into how studies are already being conducted.
What I did not hear from sites was a desire for more technology. What I heard was a desire for less administrative burden, greater clarity, and more time to focus on patients and study delivery.
One message came through clearly: sites are not looking for another platform—they are looking for help managing growing complexity.
If these tools can reduce administrative burden, simplify routine tasks, and give site teams more time to focus on patients and study delivery, they will find a place in clinical research. If they create additional work, complexity, or uncertainty, adoption will stall regardless of how advanced the technology becomes.
What gives me confidence is that sites are not resisting this conversation. Many are actively exploring where these tools may help. What they are looking for is clarity—where these tools add value, where they introduce risk, and what guardrails need to be in place.
At Novotech, we understand both the pressures sites are facing and the practical considerations that come with adopting new technologies.
If these tools help sites spend more time with patients and less time managing administrative complexity, adoption will follow. If they do not, it will not.