Choice via AI

2026-02-04

AI Summary

This study addresses the problem that AI agents may misinterpret option menus in decision recommendation settings, leading to recommendations misaligned with the decision maker’s true preferences. To resolve this, the paper proposes a choice model grounded in agentic artificial intelligence, which jointly rationalizes recommendation behavior through a strict preference relation and a monotonic interpretation mechanism. The authors innovatively introduce dual monotonicity conditions to ensure full identifiability and internal consistency of preferences, while imposing an idempotence constraint to guarantee fully rational recommendations consistent with the feasible set. Under a single acyclicity assumption, the approach successfully rationalizes AI-generated recommendations, and alignment between AI behavior and the decision maker’s preferences is effectively verified through the interplay of dual monotonicity and idempotence.

Technology Category

• Knowledge Representation and Reasoning: Preferences
• Multiagent Systems: Mechanism Design
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systems

Abstract

This paper proposes a model of choice via agentic artificial intelligence (AI). A key feature is that the AI may misinterpret a menu before recommending what to choose. A single acyclicity condition guarantees that there is a monotonic interpretation and a strict preference relation that together rationalize the AI's recommendations. Since this preference is in general not unique, there is no safeguard against it misaligning with that of a decision maker. What enables the verification of such AI alignment is interpretations satisfying double monotonicity. Indeed, double monotonicity ensures full identifiability and internal consistency. But, an additional idempotence property is required to guarantee that recommendations are fully rational and remain grounded within the original feasible set.

Problem

• AI alignment
• choice theory
• preference rationalization
• menu misinterpretation
• decision-making

Innovation

• agentic AI
• choice rationalization
• double monotonicity
• preference identification
• idempotence