本課成為整合「AI Ethical Principles」、ISO 42001、NIST AI RMF 與 EU AI Act 的研究所課程目標及內涵：
### **課程目標**
1. **全面理解AI治理框架與原則**：
- 深入學習AI Ethical Principles（人工智慧倫理原則）、ISO 42001（AI治理標準）、NIST AI RMF（風險管理框架）及EU AI Act（歐盟人工智慧法案）的核心內容。
- 探討這些框架如何影響AI技術的設計、部署與應用。
2. **提升風險管理與合規能力**：
- 培養學生在高風險AI系統中進行風險評估與治理的能力。
- 強化對AI系統合規性要求的理解，尤其針對透明性、數據治理及人類監督。
3. **強化倫理與社會責任意識**：
- 探討AI應用中的倫理挑戰與責任分配。
- 理解如何在技術創新與社會影響間取得平衡。
4. **實踐跨領域整合能力**：
- 結合法律、技術、倫理與風險管理知識，為AI技術的實施與政策制定提供支持。
### **課程內涵**
#### **1. 人工智慧倫理原則（AI Ethical Principles）**
- **核心原則**：公平性、透明性、責任制、隱私保護與安全性。
- **實踐案例**：探討如何在AI系統設計與使用中落實倫理原則。
- **國際比較**：對比各國或組織（如OECD、UNESCO）的AI倫理框架。
#### **2. ISO 42001：AI治理標準**
- **標準概述**：ISO 42001的治理結構、風險管理與合規性要求。
- **應用實踐**：如何建立AI治理體系，包括風險評估、數據管理與透明性。
- **企業案例分析**：ISO 42001在不同產業的應用。
#### **3. NIST AI RMF：人工智慧風險管理框架**
- **框架核心功能**：
- **Govern（治理）**：確保AI系統的負責任管理。
- **Map（風險映射）**：識別AI系統的潛在風險。
- **Measure（風險衡量）**：量化風險對系統運行的影響。
- **Manage（風險管理）**：制定策略以減輕風險。
- **應用場景**：高風險AI系統（如醫療、交通、金融）中的風險管理實踐。
#### **4. EU AI Act：歐盟人工智慧法案**
- **核心內容**：
- 禁止的AI應用（如社會評分系統、情感推測）。
- 高風險AI系統的要求（數據治理、透明性、人類監督）。
- 低風險與最小風險系統的規範。
- **法案實施影響**：對企業、技術開發者和監管機構的影響。
- **案例研究**：法案在醫療、教育和執法領域的應用。
#### **5. 跨框架整合與比較**
- **共通點與差異**：
- 比較AI Ethical Principles、ISO 42001、NIST AI RMF與EU AI Act的核心理念與應用場景。
- **全球視角**：
- 探討各框架在全球AI治理中的角色與影響。
#### **6. 團隊專案與實踐**
- **模擬實驗**：設計符合ISO 42001與EU AI Act要求的高風險AI系統。
- **風險評估報告**：基於NIST AI RMF進行AI系統風險管理模擬。
- **政策建議**：針對AI Ethical Principles提出具體的政策建議。
This course has become the objectives and connotations of the institute that integrates "AI Ethical Principles", ISO 42001, NIST AI RMF and EU AI Act:
### **Course Target**
1. **Comprehensive understanding of AI governance framework and principles**:
- Learn in-depth the core contents of AI Ethical Principles, ISO 42001 (AI Governance Standard), NIST AI RMF (Risk Management Framework) and EU AI Act (EU Artificial Intelligence Act).
- Explore how these frameworks affect the design, deployment and application of AI technology.
2. **Improve risk management and compliance capabilities**:
- Cultivate students' ability to conduct risk assessment and governance in high-risk AI systems.
- Strengthen understanding of AI systems' compliance requirements, especially transparency, data governance and human supervision.
3. **Strengthen the arbitrary understanding of ethics and social responsibility**:
- Explore ethical challenges and responsibility allocation in AI applications.
- Understand how to balance technological innovation and social impact.
4. **Practical cross-domain integration capabilities**:
- Combining legal, technical, ethics and risk management knowledge to support the implementation and policy formulation of AI technology.
### **Connotation of course**
#### **1. AI Ethical Principles**
- **Core Principles**: Fairness, transparency, responsibility, privacy protection and security.
- **Practical Cases**: Explore how to implement ethical principles in AI system design and use.
- **International Comparison**: Comparing AI ethics frameworks for countries or organizations (such as OECD, UNESCO).
#### **2. ISO 42001: AI Governance Standards**
- **Standard Overview**: ISO 42001's governance structure, risk management and compliance requirements.
- **Application Practice**: How to establish an AI governance system, including risk assessment, data management and transparency.
- **Enterprise Case Analysis**: Application of ISO 42001 in different industries.
#### **3. NIST AI RMF: Artificial Intelligent Risk Management Framework**
- **Framework Core Functions**:
- **Govern**: Ensure the responsible management of the AI system.
- **Map (Risk Mapping)**: Identify the potential risks of AI systems.
- **Measure (Risk Measurement)**: Quantify the impact of risk on system operation.
- **Manage (Risk Management)**: Develop strategies to reduce risks.
- **Application scenario**: Risk management practice in high-risk AI systems (such as medicine, transportation, finance).
#### **4. EU AI Act: EU Artificial Intelligence Act**
- **Core content**:
- Prohibited AI applications (such as social rating systems, emotional testing).
- Requirements for high-risk AI systems (data governance, transparency, human supervision).
- Low risk and minimum risk system regulations.
- **The impact of the implementation of the bill**: The impact on enterprises, technology developers and supervisory agencies.
- **Case Study**: Application of the bill in the medical, educational and legal fields.
#### **5. Cross-frame integration and comparison**
- **Common points and differences**:
- Compare the core concepts and application scenarios of AI Ethical Principles, ISO 42001, NIST AI RMF and EU AI Act.
- **Global Vision**:
- Explore the role and impact of each framework in global AI governance.
#### **6. Team Project and Activity**
- **Simulation Experience**: Design a high-risk AI system that meets the requirements of ISO 42001 and EU AI Act.
- **Risk Assessment Report**: AI system risk management simulation based on NIST AI RMF.
- **Policy Recommendations**: Detailed policy recommendations for AI Ethical Principles.
EU AI Act
ISO 42001
NIST AI RMF
HLEG Trtustworthy AI Principles
EU AIA CT
| 評分項目 Grading Method | 配分比例 Grading percentage | 說明 Description |
|---|---|---|
| 成果展出與課堂參與成果展出與課堂參與 Results exhibition and class participation |
5 |
