What Pistoia Alliance's AI Poll Reveals About Industry Readiness And Regulatory Confidence
A conversation between Pistoia Alliance President Becky Upton, Ph.D., and Clinical Leader Executive Editor Abby Proch
A few months ago, at the Clinical Trials Technology Congress (CTTC), a panel of regulators from the U.K. MHRA, Danish Medicines Agency, and Swedish Medical Products Agency got live results from an audience poll on the adoption of AI in clinical research.
The results were clear, informative, and instantaneous — just what Pistoia Alliance intended.
In that setting, regulators got immediate insights and feedback into industry’s sentiment toward AI, as well as its biggest questions and concerns.
In this follow-up interview, Pistoia Alliance President Becky Upton, Ph.D., reviews some of those findings, including what they reveal about current barriers to adoption, the early return on investment, emerging use cases, and the growing role of patient-generated data and social media listening.
Why was it important for Pistoia Alliance to issue this poll during Clinical Trials Technology Conference, specifically at a panel convening regulatory experts?
The event brought together clinical leaders, technology teams, and regulators at a point when AI is starting to move into clinical operations, but trust and regulatory uncertainty remain major barriers.
We intended our poll to capture how industry is feeling, while the regulator panel provided the other side of the story. The panel discussions showed the agencies are not opposed to AI but are keen to understand how it is being used and ensure its adoption is safe, controlled, and explainable. The takeaway was that regulators should be engaged early, not treated as a final hurdle. We recommend aligning industry and regulators before technology adoption becomes fragmented.
To what extent were various stakeholders represented in the poll data?
The poll fielded feedback from stakeholders directly involved in the digital transformation of clinical trials, including clinical operations leaders, pharma teams, technology providers, data specialists, and regulators.
The mix of specialties is important because AI adoption cannot be accelerated by one group alone. Sponsors need practical tools, regulators need confidence in safety and explainability, and technology providers need clear expectations for validation and auditability. The poll gives a useful indication of where the clinical trials community sees both momentum and friction.
The poll data has shown that AI is beginning to deliver value in clinical development, with 42% of respondents seeing early signs of return on investment (ROI) and a further 23% expecting ROI but not yet realizing it. Do you have any insight into how companies are evaluating ROI?
We are still at an early stage in how ROI from AI is being measured in clinical development. In many cases, companies are not looking only at direct cost savings but at operational indicators that show whether trials can run faster and with less manual burden.
Some of the metrics we saw presented at the event included increased speed of protocol amendments, faster enrollment, and fewer data queries.
Respondents also said that over the next 3–5 years, AI will have the most impact on cleaning data, data analysis, and insight generation (48%) and sourcing and engaging patients (22%). How did those results confirm expectations or reveal something new about user adoption?
It makes sense that data cleaning, analysis, and insight generation are natural early use cases because they are high volume, repetitive, and already sit within digital workflows. These three use cases are also areas where the impact of AI is easier to quantify through faster processing, fewer manual steps, and quicker decision-making.
The emphasis on AI for sourcing and engaging patients is a more novel use case. It suggests companies are starting to look beyond using AI for back-office efficiency and toward more strategic clinical trial challenges. Patient recruitment and engagement remain persistent bottlenecks, so it is encouraging to see clinical teams recognize AI’s potential to support patient-centered trial design and delivery.
The poll also found that 60% of respondents are already using, piloting, or exploring patient-generated data to inform clinical development decisions beyond marketing. To what extent is social media listening a part of that? And is there any data, here or otherwise, to understand how much of industry is leveraging social media listening?
Social media listening is only one part of the shift toward using patient-generated data to inform clinical development. Other sources include digital health tools, patient-reported outcomes, and patient support programs. The poll shows the industry is increasingly looking beyond traditional trial data because the data may not capture the full patient experience outside the trial environment. More than half (58%) of respondents said the primary benefit of social media listening is better understanding patient needs, monitoring sentiment, and tracking patient experience.
There is a huge opportunity here: AI makes it possible to analyze volumes of unstructured patient data that would have been impossible to process manually just a few years ago. But as with any new technology or process change, standards are needed. This is especially true in clinical development, where patient privacy is a major concern and decisions can directly affect patients.
The Alliance is already creating new standards in the clinical space, including through its best-practice framework for ethical social media use in drug development, which Clinical Leader recently discussed with Thierry Escudier and Aditya Tyagi. The Alliance will continue expanding its clinical work through the next phase of its social media project. The development marks our first direct patient engagement research in oncology, rare disease, and cardiology, alongside POMELO (Protocol for Social Media Listening Online), a decision-support questionnaire to help companies assess whether and how social media listening should be used.
How can readers view the complete poll results?
Readers can see the full poll results by visiting the Resource Library on our website.
How will these poll results be used within Pistoia Alliance or elsewhere to better understand and/or expand AI use in clinical research?
The poll will help to direct our future projects and communities, as well as the discussions we have with members. Many of the challenges raised at CTTC are not unique to clinical trials. They mirror issues we see in early R&D, including concerns around trust, data quality, interoperability, and skill development. That tells us the industry needs to think about technological change across the full pharma pipeline, rather than treating discovery, development, and regulatory processes as separate digital transformation challenges.
About The Author:
Becky Upton, Ph.D. was appointed as the Pistoia Alliance’s first female president in June 2022. She is a longtime supporter of pre-competitive collaboration in life sciences and healthcare R&D and the critical role it plays in advancing science and is passionate about diversity in STEM. Upton is responsible for leading the Pistoia Alliance’s strategy and defining its future within areas of increasing importance to the industry, such as data standards, emerging technologies, diversity and inclusion, sustainability, and precision medicine. Upton has a Ph.D. in biochemistry from Imperial College and an MBA from Cranfield University.