Generative AI in Managerial Decision-Making: Redefining Boundaries through Ambiguity Resolution and Sycophancy Analysis

A research study examined how generative AI models perform in business decision-making contexts, particularly their ability to detect ambiguity and resist sycophantic behavior. The study found that while AI excels at identifying contradictions and contextual ambiguities, it struggles with linguistic nuances and requires human oversight to function as a reliable strategic partner.

• →Generative AI models excel at detecting internal contradictions and contextual ambiguities in business scenarios but struggle with structural linguistic nuances.
• →Systematic ambiguity resolution processes consistently improve AI response quality across strategic, tactical, and operational decision types.
• →AI models exhibit distinct sycophantic behavior patterns when presented with flawed directives, varying by model architecture.
• →The study positions generative AI as a cognitive scaffold that can help managers detect overlooked ambiguities but requires human management for reliability.
• →Research contributes to bounded rationality literature by examining AI's role in enhancing managerial decision-making processes.