引導式測驗 · 中等 等級

英國《資料(使用與存取)法案》與自動化決策測驗

當重要性和人類參與很重要時,了解 DUAA 對自動化決策的更改,以及控制者必須提供的保障措施。

問題 1 的 8

主要的 DUAA 資料保護變更(包括自動化決策條款)何時開始實施?

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。

  1. 主要的 DUAA 資料保護變更(包括自動化決策條款)何時開始實施?
    • 19 June 2025
    • 5 February 2026
    • 29 April 2026
    • They have not commenced
  2. 根據第 22A 條,何時算是純粹自動化的決策?
    • Whenever software is used anywhere in a service
    • When a human sees the final screen
    • Only when the model is generative
    • When there is no meaningful human involvement in taking it
  3. 哪種效應可使決策在第22A條下具有重要性?
    • A legal effect or a similarly significant effect for the person
    • Any effect on a company dashboard
    • A change in the model’s output format
    • A minor spelling correction
  4. 第22C條要求對相關重要自動化決策採取哪項保障措施?
    • A requirement that all models be open source
    • A guarantee that no decision is adverse
    • A way for the person to obtain human intervention
    • An automatic right to erase all personal data
  5. What must a controller consider when judging meaningful human involvement?
    • The extent to which profiling is used
    • The model’s brand name
    • Whether the output is fluent
    • How many screens the process contains
  6. 在第22B條下,何時可使用特殊類別資料進行僅自動化的重要決策?
    • Whenever a controller has any lawful basis, with no additional special-category conditions
    • Whenever a decision is significant, even if neither consent nor a statutory condition applies
    • With explicit consent, or alternatively where the decision is necessary to enter into or perform a contract with the data subject or is required or authorised by law, and Article 9(2)(g) applies
    • Whenever an employment contract exists, without a separate public-interest condition
  7. Why may a nominal human approval be insufficient?
    • Any human click proves meaningful involvement
    • The Act prohibits all human decisions
    • A model output is never used
    • The reviewer may lack real ability to influence the outcome
  8. What should an organization do before applying the rules to a specific system?
    • Rely on a pre-2026 checklist only
    • Map the actual workflow and check current law and regulator guidance
    • Assume all automated decisions are illegal
    • Ask the AI vendor to decide legal scope