AI 服務可用性單點故障
Operation & Monitoring
Risk Description
Organizations over-rely on a single AI service provider (OpenAI, Google, Anthropic). When the provider experiences service disruptions, API quota adjustments, or pricing changes, all organizational AI-dependent functions cease. Lacking AI service redundancy and degradation strategies exposes organizations to third-party service availability risks.
Framework Mappings
iso 23894R7
tw principle資安與安全
tw principle問責
tw risk type部署互動問題（Deployment Interaction）
Controls English translation pending
- 多供應商備援策略
- 降級服務設計
Implementation Steps English translation pending
- 建立多 AI 供應商備援架構，確保主要供應商故障時可自動切換
- 設計優雅降級策略，在 AI 服務不可用時提供基本功能
- 定期進行 AI 服務中斷演練，測試備援機制
- 評估並維護關鍵 AI 功能的本地部署替代方案