Who's Sorry Now: User Preferences Among Rote, Empathic, and Explanatory Apologies from LLM Chatbots

TL;DR

This study investigates user preferences for different apology styles from LLM chatbots across various error types, revealing that explanatory apologies are generally preferred but context-dependent, highlighting the complexity of repairing trust in AI systems.

Contribution

It provides empirical insights into user preferences for apology types in LLM chatbots and emphasizes the importance of personalization and context-awareness for effective trust repair.

Findings

Explanatory apologies are generally preferred by users.

User preferences vary by error type and context.

Empathic apologies are favored for emotional acknowledgment in bias scenarios.

Abstract

As chatbots driven by large language models (LLMs) are increasingly deployed in everyday contexts, their ability to recover from errors through effective apologies is critical to maintaining user trust and satisfaction. In a preregistered study with Prolific workers (N=162), we examine user preferences for three types of apologies (rote, explanatory, and empathic) issued in response to three categories of common LLM mistakes (bias, unfounded fabrication, and factual errors). We designed a pairwise experiment in which participants evaluated chatbot responses consisting of an initial error, a subsequent apology, and a resolution. Explanatory apologies were generally preferred, but this varied by context and user. In the bias scenario, empathic apologies were favored for acknowledging emotional impact, while hallucinations, though seen as serious, elicited no clear preference, reflecting user uncertainty. Our findings show the complexity of effective apology in AI systems. We discuss key insights such as personalization and calibration that future systems must navigate to meaningfully repair trust.