Unblinded: A Clinical Trial Podcast Biggest Risks (and Possible Rewards) of AI in Clinical Data
12 snips
Sep 15, 2025 In this insightful discussion, Doug Bain, founder of ClinFlow, Drew Garty, CTO at Veeva, and Robert Bergann, leader of clinical digital innovation at Bayer, delve into the transformative power of AI in clinical development. They share practical AI use cases like pattern detection and the innovation of automated study builds. The guests also discuss the importance of human oversight in AI outputs, effective governance for AI strategies, and the need for proactive detection methods. Their call to action emphasizes focusing AI on patient benefits and quality data.
AI Snips
Chapters
Transcript
Episode notes
Use AI To Detect Unseen Patterns
- Leverage existing trial databases to detect unusual patterns that humans cannot review at scale.
- Use AI to flag anomalies like audit-trail irregularities and repetitive queries for human follow-up.
Mature Architecture Before Automation
- Treat AI as an additional detection methodology, not a drop-in replacement for deterministic logic.
- Build mature architectures and proof-of-concepts before using AI in regulated, decision-making workflows.
Accept Non‑Static AI Outputs
- Humans must monitor AI outputs to understand changing answers and evolving models.
- Expect non-static results and require mechanisms to explain why model outputs change over time.