---
id: 20260510-T0-07
title: "企业级AI智能体检索受限？新基准测试揭示授权证据缺失问题"
title_en: "New Benchmark Tests Authorization-Limited Evidence in Enterprise AI Agents"
url: https://ai.daily.yangsir.net/daily/20260510-T0-07
issue_date: 2026-05-10
publish_date: 2026-05-09T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2605.05379
---

# 企业级AI智能体检索受限？新基准测试揭示授权证据缺失问题

arXiv发表新论文《Partial Evidence Bench》。研究发现，在企业环境中，AI智能体往往受限于访问控制和策略约束，导致检索到的证据不完整，但仍会生成看似合理的错误答案。该基准测试专门评估智能体在授权受限环境下的表现，为企业部署AI提供了新的安全评估标准。

## English Version

**New Benchmark Tests Authorization-Limited Evidence in Enterprise AI Agents**

A new arXiv paper, 'Partial Evidence Bench,' reveals that enterprise agents often operate under access controls and policy constraints that limit their retrieval capabilities. This can result in incomplete evidence, yet the system still produces plausible but incorrect answers. This benchmark specifically evaluates agent performance in authorization-limited environments, offering a new safety evaluation standard for enterprise AI deployment.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2605.05379)

**详情页**：https://ai.daily.yangsir.net/daily/20260510-T0-07

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*