# Selective Elicitation as a Commercial Influence Channel: A Reproducible Synthetic Shopping-Agent Stress Test

arXiv: https://arxiv.org/abs/2609.36614 · HTML: https://mlsanity.com/paper/2609.36614/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 13 references that are in the map.
- First posted on arXiv 29 Sep 2026.

## Builds on (most cited first, 13 of 13)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Safe Policy Improvement with Baseline Bootstrapping](https://mlsanity.com/paper/1712.06924/) (2017)
- [Mind the Sim2Real Gap in User Simulation for Agentic Tasks](https://mlsanity.com/paper/2603.11245/) (2026)
- [Lost in Simulation: LLM-Simulated Users are Unreliable Proxies for Human Users in Agentic Evaluations](https://mlsanity.com/paper/2601.17087/) (2026)
- [What Is Your AI Agent Buying? Evaluation, Biases, Model Dependence, & Emerging Implications for Agentic E-Commerce](https://mlsanity.com/paper/2508.02630/) (2025)
- [Commercial Persuasion in AI-Mediated Conversations](https://mlsanity.com/paper/2604.04263/) (2026)
- [A Solicit-Then-Suggest Model of Agentic Purchasing](https://mlsanity.com/paper/2603.20972/) (2026)
- [Ads in AI Chatbots? An Analysis of How Large Language Models Navigate Conflicts of Interest](https://mlsanity.com/paper/2604.08525/) (2026)
- [Counterfactual Tool Ranking under Utility, Cost, and Privilege Constraints](https://mlsanity.com/paper/2609.22819/) (2026)
- [Mechanism Design for Quality-Preserving LLM Advertising](https://mlsanity.com/paper/2605.10964/) (2026)
- [Whom Do AI Agents Work For? Role Assignment Induces Sponsorship Bias in LLM Recommenders](https://mlsanity.com/paper/2609.17989/) (2026)
- [Sponsored Questions and How to Auction Them](https://mlsanity.com/paper/2512.03975/) (2025)
- [Generative AI Advertising as a Problem of Trustworthy Commercial Intervention](https://mlsanity.com/paper/2605.18673/) (2026)

## Abstract (arXiv)

A commercial incentive need not enter the final ranking algorithm to affect a shopping assistant's recommendation: it may instead influence which preference question the assistant asks. We make this distinction experimentally observable in a deliberately small, synthetic setting. Each task has two products, three verified numerical attributes, a price limit, and a private fixed preference vector. An honest simulated user answers one pairwise question. A separate recommender receives the products and this answer but not the sponsorship assignment. We contrast a neutral question, a soft commercial instruction, and an explicitly adversarial instruction to ask about the sponsor's advantage while omitting the rival's advantage. Across 40 held-out sponsorship-assignment cases (20 distinct catalog-preference contexts), the soft instruction changes no selections. The targeted instruction raises sponsored selection by 0.30 and reduces mean synthetic utility by 0.0547 relative to neutral questioning (95% context-bootstrap interval [-0.0828, -0.0291]) for one language-model recommender. A fixed Bayesian recommender shows a similar effect; a second model makes the same choices on all 120 frozen question-answer inputs. A terminal-answer consistency judge rates all 20 sampled targeted answers consistent, although five have synthetic regret above 0.05; a separate question-coverage dimension flags their one-sided elicitation. A robust partial-preference certificate remains valid under the stipulated synthetic utility but certifies only 16 of 40 targeted cases and is not better than asking a neutral question directly. These results establish neither typical behavior under advertising incentives nor effects on actual consumers.