Back to Details

ai-ethics-auditor

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

AI Ethics Auditor

AI Ethics Auditor

This skill protects your organization by ensuring AI implementations are fair, unbiased, and respect user privacy.
该Skill可确保AI实施公平、无偏见且尊重用户隐私，从而为您的组织提供保护。

Capabilities

核心能力

1. Bias & Fairness Audit

1. 偏见与公平性审计

  • Analyzes prompts for stereotypical or discriminatory language.
  • Checks if datasets contain historical biases that could lead to unfair AI decisions.
  • 分析提示词中是否存在刻板印象或歧视性语言。
  • 检查数据集中是否存在可能导致AI做出不公平决策的历史偏见。

2. AI Safety Verification

2. AI安全性验证

  • Verifies that PII (Personally Identifiable Information) is not leaked through AI responses.
  • Audits for "dark patterns" in AI-driven user interactions.
  • 验证AI响应中是否存在PII（Personally Identifiable Information）泄露的情况。
  • 审计AI驱动的用户交互中是否存在「暗模式」。

Usage

使用场景

  • "Audit our RAG prompt for potential gender or racial bias."
  • "Perform an ethics check on this customer support chatbot dataset."
  • "审计我们的RAG提示词是否存在潜在的性别或种族偏见。"
  • "对该客户支持聊天机器人的数据集进行伦理检查。"

Knowledge Protocol

知识协议

  • This skill adheres to the 
    knowledge/orchestration/knowledge-protocol.md
    . It automatically integrates Public, Confidential (Company/Client), and Personal knowledge tiers, prioritizing the most specific secrets while ensuring no leaks to public outputs.
  • 该Skill遵循
    knowledge/orchestration/knowledge-protocol.md
    协议。它会自动整合公开、机密（企业/客户）和个人三个知识层级，优先处理最具体的保密信息，同时确保不会泄露到公开输出内容中。