The AI Drug Discovery Audit: What Recursion's Platform Bet Means for Every Pharma R&D Leader Committing Budget to an Algorithm's Hypothesis
Recursion acquired Exscientia in 2024, absorbing Sanofi's $1.2B AI drug discovery deal. An AI-designed drug completed Phase II. FDA reviewers are asking about AI methodology in pre-IND meetings. The governance gap: pharma R&D leaders are committing Phase II budgets to AI confidence scores without a protocol for validating what the algorithm actually claims — or for documenting it when FDA asks. This episode dissects the training data bias problem, the FDA documentation gap, what happens when your platform partner is acquired, and the four governance postures every pharma R&D function needs to evaluate before the next Phase II allocation meeting.
The Deployment Debrief · Host: Elise · AI Insight Lab
Key takeaways
- 1
FDA reviewers are asking about AI methodology in pre-IND meetings — the documentation gap that exists at most pharma R&D functions is now a regulatory event, not just an internal governance question.
- 2
Training data bias in drug discovery AI affects Phase II success rates — models trained on limited genetic datasets underperform on populations not represented in training.
- 3
When your AI drug discovery platform partner is acquired, the model, training data, and methodology you validated may change without a corresponding governance event.
- 4
The governance program that passes FDA review documents training data provenance, validation methodology, and human scientist challenge protocol — not just platform vendor certifications.
Episode sections
Why FDA asking about AI methodology in pre-IND meetings is the governance event that changes the calculus for every pharma R&D function using AI.
What Recursion, Insilico Medicine, and similar platforms do in the discovery pipeline and where the AI confidence score enters the R&D decision.
Why models trained predominantly on Western European genetic datasets perform differently on other populations — and what that means for Phase II success rates.
What FDA reviewers are asking for in pre-IND meetings that most pharma R&D functions cannot currently produce about their AI-generated hypotheses.
What the Recursion-Exscientia acquisition means for validated models, training data provenance, and the methodology your team relied on.
The difference between validation experiments that confirm what the AI predicted and challenge experiments that can actually falsify the hypothesis.
Vendor-reliant, documentation-first, challenge-experiment protocol, and full AI audit program — what each requires and which your organization needs.
Phase II budget commitment to stale models, FDA documentation failure, training data liability, platform acquisition disruption, and regulatory approval delay from AI methodology questions.