研究：购物 AI 代理会像人类一样被定价线索误导
Shopping by algorithm: How agentic AI deploys human heuristics as a surrogate consumer
8 个商用 LLM 当购物代理，一旦要花钱查信息就学人类贪便宜，研究提示词写法能救回来，做电商的都该看看。
arXiv 论文提出 Tool-Lab 方法，把商品属性藏在需要付费的工具调用之后，测试营销定价线索（尾数定价、促销框架）对 AI 购物代理的影响。实验覆盖三家供应商的 8 个商用 LLM。零成本时定价线索很少误导模型；但在模糊目标提示下引入信息获取成本，LLM 会省略计算单价所需的关键属性，做出类似人类启发式决策的次优选择。相比之下，具体的目标提示能基本保持诊断性搜索和选择最优性，说明漏洞来自店面信息架构而非 LLM 本身。
Shopping by algorithm: How agentic AI deploys human heuristics as a surrogate consumer
Consumers increasingly delegate purchasing decisions to Large Language Models (LLMs) acting as surrogate consumers. Using "Tool-Lab," an adaptation of information-board process tracing that places product attributes behind costly tool calls, we examine how marketing pricing cues (i.e., just-below pricing and promotional framing) influence AI shopping agents. Across eight commercially deployed LLMs from three providers, we trace pre-choice information acquisition. Under zero cost, pricing cues rarely mislead. Imposing acquisition costs under a vague goal prompt leads LLMs to omit diagnostic attributes required to compute unit price and choose suboptimal choices resembling human heuristics. Relative to a specific goal prompt that mainly preserves diagnostic search and choice optimality, a vague goal prompt under constraints creates a search-mediated vulnerability. This research demonstrates that marketing heuristics in delegated AI shopping are governed by storefront information architecture, not necessarily immutable LLM flaws.