Generative AI in Medical Affairs: Real Use Cases from Merck, Pfizer & Astellas

Following a standing-room-only session at the 2025 MAPS Annual Meeting, Scott Thompson sits down with three industry leaders who are shaping how generative AI is being applied inside Medical Affairs today -- not in theory, but in practice.

Wendy Fraser (Merck), Annamaria Crisan (Pfizer), and Rishi Ohri (Astellas Pharma) share real-world examples, critical lessons, and hard-won truths from piloting Gen AI across Field Medical teams, Medical Information, HEOR, and Omnichannel. They discuss personal productivity, process improvement, clinical risks, compliance pitfalls, and how to build trust -- both with your technology and your teams. If you're exploring how to scale Gen AI within a regulated environment, this is your roadmap.

Key Takeaways:

• Start with Low-Risk, High-Impact Use Cases -- Early wins build organizational confidence and create the foundation for broader adoption.
• Human-in-the-Loop Is Essential for Quality and Trust -- AI outputs in regulated environments require consistent human review to maintain scientific integrity.
• Prompt Engineering Is a Leadership Priority -- How you ask the question shapes the answer; Medical Affairs leaders need to develop this skill intentionally.
• Compliance Buy-In Starts with Demonstrating Quality -- Showing legal and compliance teams real outputs builds the trust needed to move forward.
• Modular, Fit-for-Purpose AI Beats One Size Fits All -- The most effective implementations are scoped to specific workflows, not deployed as blanket solutions.

Connect with the guests on LinkedIn:

Wendy Fraser: https://www.linkedin.com/in/wendy-fraser-86316710

Rishi Ohri: https://www.linkedin.com/in/rishiohri

Annamaria Crisan: https://www.linkedin.com/in/annamaria-crisan-b0018a17

Find more episodes of OnPoint: The Medical Excellence Podcast on our OnPoint YouTube channel.

[00:00] Why Gen AI? Why Now?
[04:00] Defining the Problems to Solve
[08:00] Three Buckets of Use Cases
[12:00] Real Examples: What’s Working (and What’s Not)
[28:00] Risks, Trust & Compliance in a Regulated World
[36:00] Scaling Gen AI Without Breaking Workflow
[47:00] Upskilling & What Leaders Are Learning Next