Meeting Minutes
Artificial Intelligence in Clinical Development and Clinical Operations
Meeting Date:
July 22, 2026
Meeting Type:
CRSC Meeting
Prepared By:
Lisa Mulder (CHAIR)
Meeting Objective
The Steering Committee met to discuss practical applications of artificial intelligence (AI) across clinical development, focusing on adoption strategies, medical writing, regulatory operations, quality management, TMF, clinical business operations, and practical use cases that can improve efficiency within small biotech companies.
AGENDA / KEY QUESTIONS
01
02
What has your experience been with these models in medical/regulatory writing e.g., quality, efficiency, compliance, and overall value?
03
What other use cases have you found to deliver the most impact/value?
KEY TAKEAWAYS
Protocol Development
The committee discussed the challenges of gaining organizational acceptance of AI and shared positive experience beta-testing an AI platform for protocol development.
Protocol Development
- The committee discussed the challenges of gaining organizational acceptance of AI and shared positive experience beta-testing an AI platform for protocol development.
- The platform produced well-structured protocol drafts, improved study feasibility assessments, and operated within a secure closed environment where company data remained private.
- Document quality improved significantly as the AI platform learned the organization’s writing style and preferences.
CSR Writing
Case study compared internal CSR writing with AI-platform CSR writing. The AI platform generated a draft CSR in 3 days compared with a CRO estimate of 52 days and 7x the cost.
CSR Writing
- The committee discussed a cautious approach to selecting AI vendors and a preference for closed systems.
- A case study compared internal CSR writing with AI-platform CSR writing. The AI platform generated a draft CSR in 3 days compared with a CRO estimate of 52 days and 7x the cost.
- The draft was approximately 90% complete, supporting the approach of using AI to produce the initial draft followed by human scientific review.
- A less successful pilot was also discussed. Because the CRO did not share its CSR template with the AI vendor for review prior to starting the work, the AI-generated CSR required significant rework and struggled to accurately interpret and communicate the clinical story despite having all source documents.
Other Applications
CRSC members discussed additional AI applications including TMF reconciliation (Veeva Falcon AI), NDA preparation, informed consent simplification, site feasibility, site contracting, and vendor proposal comparisons.
Other Applications
- CRSC members discussed additional AI applications including TMF reconciliation (Veeva Falcon AI), NDA preparation, informed consent simplification, site feasibility, site contracting, and vendor proposal comparisons.
- Participants agreed that medical writing and operational document generation currently represent some of the strongest use cases while emphasizing that human oversight remains essential.
- AI was used to create a list of initial SOPs needed for a biotech; however, the output was not strong.
- AI was also used to draft a specific SOP and produced a basic form, which was still considered better than starting from a blank page.
- The committee raised the importance of vendors understanding clinical development workflows, not simply AI technology. It was also noted that several AI vendors now employ experienced medical writers and subject matter experts.
- Business operations use cases discussed included AI-assisted site contracting, automated site-payment invoice generation based on EDC data, MSA negotiation support, CRO proposal comparisons against RFIs, vendor selection, and operational analytics that reduce hours of manual work.
- AI was used to compare vendor proposals and was able to identify the main reasons for differences in pricing.
- AI was also used for drug supply projections on a small study, with positive results.
Closing Discussion
The committee encouraged continued sharing of real-world experiences and recommended evaluating AI as a productivity accelerator rather than a replacement for experienced clinical professionals.
Closing Discussion
Members agreed that virtually all major CROs are incorporating AI into internal workflows. The committee encouraged continued sharing of real-world experiences and recommended evaluating AI as a productivity accelerator rather than a replacement for experienced clinical professionals.
Key Takeaways
Medical writing remains one of the most mature AI applications in clinical development. Closed AI environments can effectively address confidentiality concerns.
Key Takeaways
- Medical writing remains one of the most mature AI applications in clinical development.
- Closed AI environments can effectively address confidentiality concerns.
- AI significantly accelerates first-draft generation and document review.
- Human oversight remains essential for scientific interpretation and regulatory quality.
- Small biotech organizations can gain substantial efficiencies through targeted AI adoption.
- Vendors Discussed: PeerAI, BioRCE, Claude, ChatGPT, Lodestar, Veeva Falcon AI, Clin AI, Condor, PhaseV Trials and Pharma Acuity.
Action Items
Continue evaluating closed AI platforms including Claude and Copilot. Monitor Veeva Falcon AI developments.
Action Items
- Continue evaluating closed AI platforms including Claude and Copilot.
- Monitor Veeva Falcon AI developments.
- Share additional experience with PRAI, BioRCE, Clin AI and other emerging vendors.
- Continue discussing practical AI use cases during future Steering Committee meetings.