Wharton Research: 80% Follow AI's Misguided Advice

A recent Wharton study has exposed a troubling pattern: 80% of professionals follow incorrect AI advice, even when they possess the expertise to know better. This phenomenon, termed cognitive surrender, poses a genuine risk to AEC teams relying on AI for estimates, risk assessments, and design reviews. The research underscores a critical gap between AI's capabilities and the human judgement required in complex project environments.

Cognitive surrender occurs when professionals defer to AI outputs without applying their own critical analysis. The Wharton research, which surveyed 1,372 individuals, found that participants consistently followed AI-generated suggestions regardless of accuracy. This is particularly concerning in AEC, where precision directly impacts project outcomes, safety, and costs.

The study builds on cognitive science frameworks that distinguish between intuitive thinking and deliberate reasoning. In project delivery, this translates to a real risk: engineers and managers may accept AI recommendations for structural analysis, cost estimation, or scheduling without the scrutiny these decisions warrant. When AI systems present data with apparent confidence, professionals often treat that confidence as validation, bypassing their own expertise.