When Guardrails Look Effective: Construct Validity Failures in LLM Agent Commerce Evaluation
Peiying Zhu, Sidi Chang
Abstract
Interactive simulations increasingly evaluate policies in markets populated by language-model agents. Their outputs can look economic---prices, profits, consumer surplus, and welfare---without instantiating the behavior named in the claim. We audit this risk in a multi-turn buyer--seller testbed for configurable hotel transactions. An initial implementation reported welfare gains from two marketplace guardrails of +87.4, +35.0, and +28.8 across a Qwen2.5 1.5B--14B ladder. It also gave guarded and unguarded agents different offer schemas and choice procedures. Holding the schema and buyer chooser fixed changes the paired contrasts to +7.2, -13.9, and +23.8. The four largest 14B single-generation effects averaged +229; after three generations per profile-condition, they averaged +37.6 (95% bootstrap interval [-34.2, 109.3]), while generation residuals account for 49.9% of variation in this post-hoc probe. A seller-incentive check is non-monotone: increasing profit pressure produces less profit than the default seller prompt. Scripted positive controls show why this matters. A profit-maximizing seller already attains first-best welfare, so guardrails mostly redistribute and reduce welfare; they create welfare only when the seller is explicitly programmed to force inefficient bundles. We contribute a construct-validity contract separating incentive validity, protocol isolation, stochastic stability, and welfare accounting, and returning INVALID or INCONCLUSIVE before substantive policy claims. In our case, the original estimate is INVALID under protocol isolation, while the controlled study remains INCONCLUSIVE under incentive validity and stochastic stability. The case does not show that guardrails are ineffective; it shows their apparent value is unidentified until the simulated agents and protocol pass these checks.
Create a lesson
Related papers
Discriminative World Models for Web Agents
Kelvin Li, Dhruv Pendharkar, Anish Pahilajani et al.
AI Contextual Measurement for Recovering Individual and Group-Level Effects: Validation Against Survey Measures and an Occupational Application
Wenxin Jiang, Xuyang Wang, Yuxiao Wu
Large Language Models (LLMs) for Telecom Root Cause Analysis (RCA): A Structured Reasoning Framework for Evidence-Grounded Diagnosis
Hao Zhou, Mandar Kulkarni, Hao Chen et al.
SafeEvolve: Harness-Policy Co-Evolution from Agent Experience for Safety Alignment
Qinghua Mao, Wanying Qu, Dadi Guo et al.
Measurement-Driven Sub-Network Selection for On-Premise Retrieval-Augmented Factory Agents
Vasileios Rizeakos, Georgios Paisios, Alexandros Machairas et al.
Bilevel Coordinated Reflection: A Game-Theoretic Approach to Multi-Agent LLM Systems
Yihang Chen, Yuxiang Chen, Yuxuan Huang et al.