引導式測驗 · 中等 等級
中國的PIPL與自動決策測驗
審查 PIPL 自動決策要求、行銷選擇、重大決策權、合法依據、影響評估和敏感資料限制。
問題 1 的 8
What does PIPL Article 24 require for automated decision-making?
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。
- What does PIPL Article 24 require for automated decision-making?
- Transparency and fair, impartial results
- Automatic consent for every model
- Public release of source code
- A ban on all automated tools
- What must personalized information-push or commercial marketing provide?
- A right to delete the provider
- An option to view model weights
- A non-personalized option or a convenient way to refuse
- A guarantee of no advertising
- When can a person request an explanation and refuse a solely automated decision?
- For every low-impact product recommendation
- When the decision may significantly affect their rights or interests
- Only if the provider is outside China
- Only when no personal information is used
- Does PIPL Article 24 ban all automated decisions?
- Yes, all automation is forbidden
- Yes, except for shopping
- No, it applies only to hiring
- No; it imposes requirements and rights for covered decisions
- Is consent the only possible lawful basis under PIPL Article 13?
- 是的,總是
- Yes, except for public services
- No; Article 13 provides multiple lawful bases
- No, but only for public data
- Which assessment does Article 55 require before covered use of personal information for automated decision-making?
- A personal-information protection impact assessment for automated decision-making
- An EU conformity assessment
- A CAC algorithm filing for every system
- A public disclosure of training data
- Why document what a human reviewer actually examines and can change?
- It removes all privacy duties
- It makes the model open source
- It changes the provider’s country
- To assess the real decision process instead of relying on a nominal approval step
- Which other controls may apply to sensitive personal information?
- Automatic exemption from consent
- Additional protections and processing requirements under PIPL
- A blanket right to publish the data
- No extra safeguards