Behavioral Economics of AI: LLM Biases and Corrections

TL;DR

This paper investigates biases in large language models used in economic decision-making, revealing systematic patterns and demonstrating that prompting models to be rational can mitigate these biases.

Contribution

It provides the most comprehensive experimental analysis of LLM biases in economic tasks and shows how prompting can reduce these biases.

Findings

Larger and more advanced models exhibit more human-like biases in preferences.

Advanced models often produce rational responses in belief-based tasks.

Prompting LLMs to make rational decisions reduces behavioral biases.

Abstract

Do generative AI models, particularly large language models (LLMs), exhibit systematic behavioral biases in economic and financial decisions? If so, how can these biases be mitigated? Drawing on the cognitive psychology and experimental economics literatures, we conduct the most comprehensive set of experiments to date$-$originally designed to document human biases$-$on prominent LLM families across model versions and scales. We document systematic patterns in LLM behavior. In preference-based tasks, responses become more human-like as models become more advanced or larger, while in belief-based tasks, advanced large-scale models frequently generate rational responses. Prompting LLMs to make rational decisions reduces biases.