Cognitive Biases in LLM-Assisted Software Development

2026-01-12

AI Summary

This study systematically investigates the mechanisms, types, and impacts of cognitive biases in large language model (LLM)-assisted software development. Employing a mixed-methods approach—including observational experiments with 14 developers, a survey of 22 participants, qualitative coding, and validation by psychologists—the research establishes the first comprehensive taxonomy of cognitive biases in this context, comprising 15 categories encompassing 90 specific bias instances. The findings reveal that 48.8% of developer behaviors are influenced by cognitive biases, with 56.4% of these cases directly attributable to human–LLM interaction. This taxonomy provides a theoretical foundation and actionable mitigation strategies for the design of LLM-augmented programming tools and development practices.

Technology Category

• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

The widespread adoption of Large Language Models (LLMs) in software development is transforming programming from a solution-generative to a solution-evaluative activity. This shift opens a pathway for new cognitive challenges that amplify existing decision-making biases or create entirely novel ones. One such type of challenge stems from cognitive biases, which are thinking patterns that lead people away from logical reasoning and result in sub-optimal decisions. How do cognitive biases manifest and impact decision-making in emerging AI-collaborative development? This paper presents the first comprehensive study of cognitive biases in LLM-assisted development. We employ a mixed-methods approach, combining observational studies with 14 student and professional developers, followed by surveys with 22 additional developers. We qualitatively compare categories of biases affecting developers against the traditional non-LLM workflows. Our findings suggest that LLM-related actions are more likely to be associated with novel biases. Through a systematic analysis of 90 cognitive biases specific to developer-LLM interactions, we develop a taxonomy of 15 bias categories validated by cognitive psychologists. We found that 48.8% of total programmer actions are biased, and developer-LLM interactions account for 56.4% of these biased actions. We discuss how these bias categories manifest, present tools and practices for developers, and recommendations for LLM tool builders to help mitigate cognitive biases in human-AI programming.

Problem

• cognitive biases
• LLM-assisted software development
• developer decision-making
• human-AI collaboration
• software engineering

Innovation

• cognitive biases
• LLM-assisted development
• human-AI collaboration
• developer decision-making
• bias taxonomy