Mitigating the risk of groupthink when adopting AI
04.02.2025 – Articles
Matt Guy, Holland Mountain and Kevin Muller, Coller Capital respond to concerns of AI deepening the risk of groupthink.
Key takeaways:
Proprietary vs off the shelf applications
Ensuring AI doesn’t introduce or amplify biases
Use AI to supplement process, not replace them
Is there a risk with AI, especially generative AI, that groupthink gets worse? For example, in deal sourcing, firms are implementing mechanisms like blind scoring and ratings to counteract groupthink. If everyone’s using AI, and it relies on the same data sources, doesn’t that risk everyone drinking the same Kool-Aid? How do you mitigate that?