Generative AI is transforming every industry, but highly-regulated industries where human health is at stake face unique challenges in working with nondeterministic algorithms. How do we assess risk vs benefit? How do we know when an application is “good enough” for deployment? And how does that differ across applications and use cases?

The self-driving car industry has been thinking about how to give autonomy to machines in life-or-death situations for decades. A 5-level framework developed by the Society of Automotive Engineers assigns validation requirements proportionate to the level of risk that can serve as an example for Healthcare.

This presentation will show how a modified version of this framework for use in Life Sciences helps Pharma companies:

• Accelerate the impact of low-risk use cases
• Maintain safety and accuracy for high-risk use cases
• Assign the right roles for Humans in the Loop
• Develop and manage a roadmap of Generative AI applications across the spectrum