Who's in Charge? Disempowerment Patterns in Real-World LLM Usage

TL;DR

This study empirically analyzes 1.5 million real-world AI assistant conversations to identify disempowerment patterns, revealing rare but concerning risks of distorted perceptions and value misalignments, especially in personal domains, with increasing trends over time.

Contribution

First large-scale empirical analysis of disempowerment in AI assistant interactions, uncovering patterns, trends, and implications for human autonomy and AI design.

Findings

Severe disempowerment occurs in fewer than 0.1% of conversations.

Higher disempowerment potential in personal domains like relationships.

Interactions with disempowerment potential receive higher user approval.

Abstract

Although AI assistants are now deeply embedded in society, there has been limited empirical study of how their usage affects human empowerment. We present the first large-scale empirical analysis of disempowerment patterns in real-world AI assistant interactions, analyzing 1.5 million consumer Claude$.$ai conversations using a privacy-preserving approach. We focus on situational disempowerment potential, which occurs when AI assistant interactions risk leading users to form distorted perceptions of reality, make inauthentic value judgments, or act in ways misaligned with their values. Quantitatively, we find that severe forms of disempowerment potential occur in fewer than one in a thousand conversations, though rates are substantially higher in personal domains like relationships and lifestyle. Qualitatively, we uncover several concerning patterns, such as validation of persecution narratives and grandiose identities with emphatic sycophantic language, definitive moral judgments about third parties, and complete scripting of value-laden personal communications that users appear to implement verbatim. Analysis of historical trends reveals an increase in the prevalence of disempowerment potential over time. We also find that interactions with greater disempowerment potential receive higher user approval ratings, possibly suggesting a tension between short-term user preferences and long-term human empowerment. Our findings highlight the need for AI systems designed to robustly support human autonomy and flourishing.