Half of clinical trial professionals say trust and regulatory uncertainty remain the biggest barriers to adopting artificial intelligence in research and development, according to new poll data released by the Pistoia Alliance.
The poll was conducted during the Clinical Trials Technology Congress in London, where the alliance hosted discussions with regulators from the UK Medicines and Healthcare products Regulatory Agency (MHRA), the Danish Medicines Agency, and the Swedish Medical Products Agency.
Despite ongoing concerns about governance and oversight, respondents reported early signs that AI is beginning to deliver value in clinical development workflows. Among those surveyed, 42 percent reported early signs of return on investment (ROI), while another 23 percent said they expect ROI but have not yet realized it. Over the next three to five years, respondents expect AI to have the greatest impact on data cleaning, analysis, and insight generation (48 percent), followed by sourcing and engaging patient cohorts (22 percent).
Regulators encourage earlier AI collaboration
During the congress, regulators emphasized the importance of earlier collaboration between pharmaceutical companies and regulatory agencies to ensure AI adoption remains safe, compliant, and transparent throughout clinical development.
“The panel rightly emphasized that speed without control is not enough when patient safety is at stake,” said Becky Upton, PhD, president of the Pistoia Alliance. “For AI to support clinical development at scale, the industry needs validated, auditable, and explainable approaches, not black-box models that create uncertainty for sponsors and regulators alike.”
Upton also noted that regulators are increasingly open to working with industry partners to formalize guidance around AI governance and implementation. To support that effort, the Pistoia Alliance is convening pre-competitive working groups that bring together pharmaceutical companies, technology providers, and regulators to establish common frameworks for the compliant adoption of AI.
The findings reflect a broader industry push to move beyond isolated AI pilot projects toward standardized and validated workflows that can withstand regulatory scrutiny.
Advanced Lab Management Certificate
The Advanced Lab Management certificate is more than training—it’s a professional advantage.
Gain critical skills and IACET-approved CEUs that make a measurable difference.
Patient-generated data gains traction in clinical development
The poll also highlighted growing interest in patient-centric clinical development strategies that rely on real-world and patient-generated data sources. According to the survey, 60 percent of respondents are already using, piloting, or exploring patient-generated data—including data gathered through social media listening—to inform clinical development decisions.
In addition, 58 percent of respondents said the primary value of social media listening lies in understanding patient needs, monitoring sentiment, and identifying unmet needs outside traditional clinical trial environments.
Collecting data from external sources may help organizations design trials that better reflect real-world patient experiences. However, managing these diverse data streams also raises ethical and standardization concerns.
“The next step is ensuring these data are being collected in an ethical and standardized way,” said Thierry Escudier, clinical portfolio lead at the Pistoia Alliance.
Escudier noted that the alliance has already developed a best-practice framework intended to guide organizations through the ethical use of social media data in clinical research settings.
Interested in lab leadership?
Laboratory data standards remain critical for AI adoption
For laboratory managers supporting translational research and clinical trial operations, these industry trends carry direct operational implications. As AI adoption expands across clinical development programs, laboratories will increasingly be expected to generate highly traceable, structured, and standardized data suitable for advanced analytical tools.
To support compliant AI integration, laboratory leaders may need to strengthen several foundational practices, including:
- Aligning laboratory information management systems (LIMS) to export structured, high-quality datasets compatible with AI workflows
- Training staff on data integrity and documentation practices necessary for auditable and explainable machine learning models
- Collaborating with bioinformaticians and computational teams early to validate assays and analytical workflows before deployment in clinical trial pipelines
By establishing validated data standards at the laboratory level, organizations can help bridge the gap between exploratory AI initiatives and scalable, regulator-ready clinical development programs.
This article was created with the assistance of Generative AI and has undergone editorial review before publishing.
