Raja Shankar, VP of machine learning at IQVIA, discusses which AI capabilities sponsors are most likely to adopt first to streamline trial workflows and reduce operational burden, while also highlighting emerging applications that could shape the next phase of clinical trial design.
In a video interview prior to the 2026 SCOPE Summit with Applied Clinical Trials, Raja Shankar, VP of machine learning at IQVIA, discussed how artificial intelligence is beginning to reshape research and development across the clinical trial lifecycle. Shankar outlined how AI-enabled trial simulation and automation are influencing protocol design, site activation, monitoring, and closeout, while also pointing to emerging applications such as synthetic control arms and digital twins that could significantly alter trial design in the years ahead.
Editor's note: This transcript is a lightly edited rendering of the original audio/video content. It may contain errors, informal language, or omissions as spoken in the original recording.
ACT: What AI solutions stand out in optimizing clinical trials with improved outcomes, streamlined workflows, scalability and reducing burdens on sponsors, sites and/or patients?
Shankar: There’s definitely a timeline to this. Some impacts will come relatively soon, especially around content generation. Informed consent form generation is one example, and that’s something sponsors will likely start using agentic AI for this year. Clinical study reports and protocol authoring are other areas where we’ll see early adoption.
Another area where change will happen relatively soon is the trial master file. Using agentic AI for TMF can increase document throughput, improve quality, and ensure the right documents are filed at the right time with appropriate guardrails. Monitoring is another major area where we’ll start to see impact, likely toward the later part of this year and into early next year.
Beyond that, we’ll begin to see early signs this year of life sciences foundation models trained on large volumes of patient and clinical data. These models support trial simulation and help sponsors predict outcomes when making decisions about trial design and patient populations. I expect that capability to ramp up more significantly in 2027 and beyond.
ACT: What are some lesser-known uses of AI in clinical trials that are noteworthy?
Shankar: One potentially disruptive area is the use of AI to create synthetic control arms and digital twins. There’s growing interest in this space, including from regulators, because these approaches could help accelerate patient access to new therapies, especially in rare diseases or high unmet-need areas where patient recruitment is challenging.
Life sciences foundation models aren’t only useful for trial simulation. They could also support the creation of synthetic control arms or digital twins, enabling trials where you have an intervention arm alongside a synthetic comparator arm that regulators may ultimately accept.
This isn’t something we’ll see broadly this year, but we may see early traction in 2027, with more widespread adoption in 2028 and beyond. If implemented successfully, these approaches could help get therapies to patients faster and potentially reduce the overall cost of clinical trials.
