When Agent Automation Becomes Profitable: Quantifying and Insuring Autonomous AI Risk through Trace-Economic Underwriting

  • 类型:arxiv
  • 标识:2606.16465
  • 链接:http://arxiv.org/abs/2606.16465v1
  • 主分类:agent
  • 形态:application
  • 被引:8
  • 被引来源:Semantic Scholar
  • S2被引:8
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:T trace-economic underwriting is introduced, which maps tool-use traces to customer exposure and claimable loss, then uses this representation for pricing, control, and risk transfer, and uses deterministic economic labels rather than an LLM judge.
  • OpenAlex ID:W7164898166
  • OpenAlex DOI:10.48550/arxiv.2606.16465
  • DOI:10.48550/arxiv.2606.16465
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2606.16465
  • OpenAlex更新:2026-07-19
  • 待LLM分类:否
  • 标题中文:当 Agent 自动化变得有利可图:通过 Trace-Economic Underwriting 量化并承保自主 AI 风险
  • TLDR中文:本文引入 trace-economic underwriting,将工具调用 trace 映射为客户风险敞口与可索赔损失,并以此表示用于定价、控制与风险转移,使用确定性经济标签而非 LLM 评判器。
  • 来源文件
  • /inbox/tom/_candidates/2026-06-17-agent-memory-tool-use-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]