引導式測驗 · 中等 等級

FTC 關於廣告測驗中人工智慧的規則

測試人工智慧能力聲明的證據、當前聯邦貿易委員會關於虛假推薦的審查規則、背書背景以及機構材料的不同狀態。

問題 1 的 8

A company says its AI fraud detector is “99.9% accurate.” What is the best first step before using the claim?

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. A company says its AI fraud detector is “99.9% accurate.” What is the best first step before using the claim?
    • Add “AI-powered” in smaller text so the number needs no evidence
    • Use an internal demo instead of representative testing
    • Assume that accuracy is implied by the product category
    • Substantiate the exact performance claim under relevant test conditions
  2. Which description of the FTC’s Consumer Reviews and Testimonials Rule is accurate?
    • It replaces all substantiation requirements for AI products
    • It prohibits specified fake or false reviews and testimonials, including those generated by AI
    • It bans every review written with any software assistance
    • It requires every advertisement to identify the model vendor
  3. A brand pays a creator to praise a product using a generative avatar. What should the advertiser examine?
    • Whether generative systems are exempt from endorsement rules
    • Whether the platform allows avatars, which settles FTC compliance
    • Whether the ad’s overall impression is truthful and any material endorsement connection is clearly disclosed
    • Whether the avatar’s face is photorealistic enough to count as a human
  4. What does the finalized DoNotPay order illustrate?
    • A specific prohibition tied to that company’s unsupported lawyer-equivalence claim
    • A blanket ban on marketing any AI product
    • A ruling that every AI chatbot must employ licensed attorneys
    • A general safe harbor for capability claims that include a disclaimer
  5. A complaint alleges a company made deceptive AI advertising claims. How should a writer describe that status?
    • As evidence that any AI use violates federal law
    • As allegations unless and until there is a final adjudication or settlement order
    • As proven facts because the FTC filed the complaint
    • As a binding rule for every advertiser
  6. Does adding a disclaimer automatically make an unsupported AI performance claim acceptable?
    • Yes, if the disclaimer is placed in the terms of service
    • Yes, whenever the product uses a real machine-learning model
    • No; disclaimers are prohibited in all AI advertising
    • No; a disclosure does not replace evidence supporting the main claim
  7. How should “AI-powered” be treated in an advertisement?
    • Assess what consumers are likely to infer and whether that implied capability is supported
    • Treat it as a purely technical label with no consumer meaning
    • Treat it as inherently deceptive in every context
    • Use it as proof that the system outperforms human experts
  8. Which is the soundest way to use FTC materials when advising a campaign?
    • Rely on a press release headline without reading the cited action
    • Apply one past order’s exact remedy to every AI product
    • Distinguish guidance, proposed or filed allegations, finalized rules, and orders by their legal status
    • Treat all FTC publications as identical binding statutes