From Personalization to Prejudice: Bias and Discrimination in Memory-Enhanced AI Agents for Recruitment

TL;DR

This paper investigates how memory-enhanced personalization in AI agents, especially in recruitment scenarios, can introduce and amplify bias, highlighting the need for safeguards.

Contribution

It is the first to systematically analyze bias in memory-enhanced personalized AI agents, demonstrating how bias is introduced and reinforced during operation.

Findings

Bias is systematically introduced during personalization.

Memory amplification reinforces existing biases.

Protective measures are necessary to mitigate bias.

Abstract

Large Language Models (LLMs) have empowered AI agents with advanced capabilities for understanding, reasoning, and interacting across diverse tasks. The addition of memory further enhances them by enabling continuity across interactions, learning from past experiences, and improving the relevance of actions and responses over time; termed as memory-enhanced personalization. Although such personalization through memory offers clear benefits, it also introduces risks of bias. While several previous studies have highlighted bias in ML and LLMs, bias due to memory-enhanced personalized agents is largely unexplored. Using recruitment as an example use case, we simulate the behavior of a memory-enhanced personalized agent, and study whether and how bias is introduced and amplified in and across various stages of operation. Our experiments on agents using safety-trained LLMs reveal that bias is systematically introduced and reinforced through personalization, emphasizing the need for additional protective measures or agent guardrails in memory-enhanced LLM-based AI agents.