引導式測驗 · 中等 等級
歐盟人工智慧法案上市後監測和嚴重事件通報測驗
測試本指南的《歐盟人工智慧法案》範圍、角色、責任、實際限制以及組織的實際部署注意事項。
問題 1 的 8
第 72 條如何支援高風險系統投入服務後的監督?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- 第 72 條如何支援高風險系統投入服務後的監督?
- Systematically collect and analyse post-market information about high-risk system performance
- Replace all user support
- Guarantee that no incident can occur
- Monitor only social-media mentions
- 第 73 條指示提供者向何處通報重大事件?
- Only to the system’s users
- Only to the model supplier
- To the market surveillance authority of the Member State where it occurred
- Only to the European Parliament
- 針對與死亡相關的重大事件,有什麼特別的截止日期?
- Within one year
- As soon as causation is established or suspected, and no later than ten days after awareness
- Only after a court ruling
- At the next annual review
- 每一項投訴都自動符合嚴重事件的資格嗎?
- Yes, all complaints must be reported as serious incidents
- No, complaints should never be retained
- Yes, if the model is generative
- No; it is a signal that requires assessment against the legal definition
- Why preserve the awareness timestamp?
- It proves technical causation
- It replaces system logs
- Reporting timelines can depend on when the provider became aware
- It determines whether an AI model is high-risk
- How should telemetry alerts be used?
- As a trigger for triage and investigation
- As automatic proof of a serious incident
- As a substitute for authority notification
- As a reason to delete logs
- What does post-market monitoring complement?
- The AI Act itself
- Every privacy obligation
- No other governance process
- Risk assessment, quality management, and deployer monitoring
- Which application dates should a team check?
- Only the general entry-into-force date
- The date for its system category and applicable obligation
- One date for every AI rule
- The vendor’s release date only