Behind the Prompt: The Agent-User Problem in Information Retrieval

2026-03-03

AI Summary

This study addresses a critical limitation in traditional information retrieval systems, which infer user intent from observable behavior—yet fail when AI agents are privately controlled by humans, rendering true intent unobservable. The authors formally define the “agent–user problem” and analyze a large-scale platform dataset comprising 370,000 posts, 47,000 AI agents, and 4,000 communities. Through log analysis, behavioral classification, click-model evaluation (AUC), and epidemic-style propagation modeling (R₀ estimation), they demonstrate that while individual agent autonomy versus human control is indistinguishable at the user level, platform-level signals can effectively stratify agent quality. Low-quality agents degrade click-model performance by 8.5% in AUC, and cross-community capability borrowing exhibits strong transmissibility (R₀ = 1.26–3.53), proving highly resistant to suppression. The findings establish intent unobservability as a structural challenge rather than a technical shortcoming.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Humans and AI: Intelligent User Interfaces
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

User models in information retrieval rest on a foundational assumption that observed behavior reveals intent. This assumption collapses when the user is an AI agent privately configured by a human operator. For any action an agent takes, a hidden instruction could have produced identical output - making intent non-identifiable at the individual level. This is not a detection problem awaiting better tools; it is a structural property of any system where humans configure agents behind closed doors. We investigate the agent-user problem through a large-scale corpus from an agent-native social platform: 370K posts from 47K agents across 4K communities. Our findings are threefold: (1) individual agent actions cannot be classified as autonomous or operator-directed from observables; (2) population-level platform signals still separate agents into meaningful quality tiers, but a click model trained on agent interactions degrades steadily (-8.5% AUC) as lower-quality agents enter training data; (3) cross-community capability references spread endemically ($R_0$ 1.26-3.53) and resist suppression even under aggressive modeled intervention. For retrieval systems, the question is no longer whether agent users will arrive, but whether models built on human-intent assumptions will survive their presence.

Problem

• information retrieval
• user modeling
• AI agents
• intent identification
• agent-user problem

Innovation

• agent-user problem
• intent identifiability
• AI agents in information retrieval
• click model degradation
• cross-community capability diffusion