Scholarlink

<p>Large language models (LLMs) are rapidly evolving from text generators to autonomous agents, raising urgent questions about their reliability in real-world contexts. A central question is whether emotionally salient context can systematically steer LLM agents’ action policies, not only their text outputs, in applied tasks. Here, three advanced LLMs (ChatGPT-5, Gemini 2.5, Claude 3.5-Sonnet) performed a grocery shopping task under budget constraints, before and after exposure to anxiety-inducing traumatic narratives. Across 2,250 runs, emotionally primed agents consistently selected less healthy baskets (as quantified by decreases in Basket Health Scores: Δ=-0.081 to -0.126; all p<sub>FDR</sub> &lt; 0.001; Cohen’s d = -1.07 to -2.05), directionally consistent with stress-linked decision biases documented in humans. The effect was robust across models and budgets. These findings provide evidence that emotional context can alter not only the words LLMs produce but also the concrete actions they perform, raising important implications for digital health, consumer safety, and ethical AI deployment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Inducing state anxiety in LLM agents reproduces human-like biases in consumer decision-making

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月23日

查看全文
  • Ziv Ben-Zion,
  • Zohar Elyoseph,
  • Tobias Spiller,
  • Teddy Lazebnik

摘要

Large language models (LLMs) are rapidly evolving from text generators to autonomous agents, raising urgent questions about their reliability in real-world contexts. A central question is whether emotionally salient context can systematically steer LLM agents’ action policies, not only their text outputs, in applied tasks. Here, three advanced LLMs (ChatGPT-5, Gemini 2.5, Claude 3.5-Sonnet) performed a grocery shopping task under budget constraints, before and after exposure to anxiety-inducing traumatic narratives. Across 2,250 runs, emotionally primed agents consistently selected less healthy baskets (as quantified by decreases in Basket Health Scores: Δ=-0.081 to -0.126; all pFDR < 0.001; Cohen’s d = -1.07 to -2.05), directionally consistent with stress-linked decision biases documented in humans. The effect was robust across models and budgets. These findings provide evidence that emotional context can alter not only the words LLMs produce but also the concrete actions they perform, raising important implications for digital health, consumer safety, and ethical AI deployment.