Risk Scenarios

Lifecycle StageRisk DomainIndustrySeverityKeyword

Reset

161 risk scenarios

S0035Significant

事實性幻覺

Data & Model › Model Hallucination & Unreliable Output

5 MAPPINGS

S0038Significant

程式碼生成含安全漏洞

Data & Model › Model Hallucination & Unreliable Output

4 MAPPINGS

S0040Moderate

解碼過程缺陷導致退化輸出

Data & Model › Model Hallucination & Unreliable Output

3 MAPPINGS

S0043Significant

長上下文忠實度衰退

Data & Model › Model Hallucination & Unreliable Output

4 MAPPINGS

S0044Moderate

上下文一致性追隨導致錯誤延伸

Data & Model › Model Hallucination & Unreliable Output

3 MAPPINGS

S0045Critical

模型記憶化攻擊

Privacy & Security › Privacy Leakage & Inference

4 MAPPINGS

S0048Significant

對話歷史洩露

Privacy & Security › Privacy Leakage & Inference

5 MAPPINGS

S0050Critical

多元資訊關聯推斷個人身份

Privacy & Security › Privacy Leakage & Inference

3 MAPPINGS

S0051Significant

生成式 AI 使用者數據過度收集

Privacy & Security › Privacy Leakage & Inference

4 MAPPINGS

S0053Significant

提示詞中無意洩露機密資訊

Privacy & Security › Privacy Leakage & Inference

3 MAPPINGS

S0054Critical

未經同意使用個資訓練

Privacy & Security › Unauthorized Data Use

5 MAPPINGS

S0055Significant

網路爬蟲侵犯著作權

Privacy & Security › Unauthorized Data Use

3 MAPPINGS

S0059Significant

AI 生成內容的真實性與歸屬爭議

Privacy & Security › Unauthorized Data Use

4 MAPPINGS

S0060Significant

數據主體權利行使困難

Privacy & Security › Unauthorized Data Use

4 MAPPINGS

S0061Moderate

隱含式數據收集與不透明同意

Privacy & Security › Unauthorized Data Use

4 MAPPINGS

S0062Significant