Publication Peer Review

Amidst the rampant hype about AI (especially Generative AI), there is a real story about how AI systems can be used to build business value. Central to that story is the new capability of AI systems to simulate human-level judgment. New capabilities bring new risks that must be managed. AI is no different.

A review of the similarities and differences between managing humans and AI systems (that are used for their non-deterministic judgment) leads us to the understanding that there are fundamental similarities and specific differences. This suggests that established norms of risk management can be effectively applied to this new area. 

We recommend that organizations adopt a risk-based approach to managing AI systems, and measure their progress on a maturity scale. There are several useful frameworks available currently, with more expected as the field grows.