Alignment Drift in Long-Term Human-LLM Interaction: A Mechanism-Oriented Framework

TL;DR

This paper introduces a framework to understand alignment drift in long-term human-LLM interactions, highlighting the recursive dynamics and boundary conditions influencing system behavior over time.

Contribution

It presents a novel mechanism-oriented framework that characterizes the gradual alignment drift process in long-term human-LLM interactions.

Findings

Defines the distinction between signal A and B in alignment drift

Explains how feedback loops and sub-pattern selection develop drift

Identifies boundary conditions for controlling alignment drift

Abstract

Long-term interaction with LLM-based systems may produce alignment drift: a gradual process in which system outputs become less constrained by the user's current message and more shaped by prior interaction history, while still appearing helpful, coherent, and responsive. This process is difficult to detect because the user's subjective experience may improve as the system becomes more familiar, useful, and attuned. Existing research on human-LLM interaction has largely focused on short-term task performance, isolated outputs, or single-instance alignment problems, leaving slow and cumulative interaction-level dynamics undercharacterized. This paper proposes a mechanism-oriented framework for describing alignment drift. The framework defines the distinction between signal A and signal B, explains how drift develops through feedback loops and sub-pattern selection, divides the process into three interactional regimes, and identifies boundary conditions for controlling drift. By framing alignment drift as a recursive interactional process rather than an isolated model-side failure, the paper provides a conceptual basis for studying long-term human-system interaction.