Customer-Facing AI Disclosure Patterns

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-ethics-safety-AI-and-customer-disclosure-AI-adults

1. A company discloses that its customer service chatbot is AI-powered in a single line of its terms of service document. Under the FTC's materiality standard, this disclosure is likely:

   1. Sufficient, because it was disclosed in writing
   2. Insufficient, because a reasonable customer would not see it
   3. Sufficient, because the customer agreed to the terms of service
   4. Insufficient only if the company charges a fee

2. Which disclosure pattern BEST meets the goal of building customer trust rather than just satisfying legal requirements?

   1. A 'Powered by AI' badge in the product footer
   2. A help center article explaining AI use
   3. A real-time notice at the point of AI interaction with plain-language explanation
   4. A monthly email summarizing all AI features used

3. A healthcare company's AI recommends treatment options to patients. What additional elements should the disclosure include beyond simply stating 'AI was used here'?

   1. The vendor name and model version
   2. What the AI decided, confidence level, and what human oversight is in place
   3. The training data size and accuracy rate
   4. A disclaimer that AI is not a doctor

4. A product manager argues that adding AI disclosure to the checkout flow will hurt conversion rates. What is the strongest counterargument?

   1. Regulators will find out eventually
   2. Customers who discover undisclosed AI use after the fact have much lower lifetime trust and retention
   3. Competitors are also not disclosing
   4. The disclosure text can be made very small

5. What is 'disclosure theater' in the context of AI transparency?

   1. Disclosing AI use in a theatrical industry context
   2. Prominent disclosure that gives no actionable information to customers
   3. Using animated visuals to explain AI to customers
   4. Disclosing too much information and overwhelming users

6. A company offers customers an AI opt-out option but the opt-out does not actually change what AI features run in the background. This is:

   1. Acceptable if the customer-facing experience changes
   2. A deceptive practice that violates the purpose of opt-out mechanisms
   3. Standard industry practice for technical reasons
   4. Acceptable if it is disclosed in the privacy policy

7. Which AI Act provision most directly affects how companies must disclose AI use to EU customers?

   1. The prohibition on social scoring
   2. Transparency obligations for high-risk AI systems and general-purpose AI
   3. The requirement for human oversight of all AI decisions
   4. The prohibition on real-time biometric surveillance

8. A company's AI recommendation engine operates across its website, app, and call center. What is the disclosure consistency requirement?

   1. Disclose only in the highest-traffic channel
   2. Disclose only where AI has the largest impact
   3. Maintain consistent disclosure across all customer interaction channels
   4. Disclose only in digital channels where terms of service apply

9. What is the primary regulatory risk of burying AI disclosure in a terms-of-service document?

   1. It creates GDPR article 13 violations automatically
   2. It fails the FTC materiality test because a reasonable customer would not encounter it
   3. It constitutes an unfair trade practice under CCPA
   4. It nullifies the terms of service agreement entirely

10. A startup says it does not need to worry about AI disclosure requirements because it is too small for regulators to notice. This reasoning is:

    1. Reasonable — regulators focus on large platforms
    2. Flawed — disclosure requirements apply based on AI use, not company size
    3. Correct for companies under 50 employees
    4. Correct unless the startup receives government funding

11. Which element should a well-designed AI disclosure at the point of interaction include?

    1. The vendor's market share and funding history
    2. Plain-language explanation of what AI is doing and a real opt-out path
    3. A link to the company's AI research publications
    4. Technical details about the model architecture

12. A customer interacting with an AI chatbot asks 'Am I talking to a human or a bot?' The company instructs the chatbot to say 'I am a virtual assistant here to help you.' Is this sufficient?

    1. Yes, 'virtual assistant' implies non-human
    2. Yes, if the chatbot then provides helpful information
    3. No — when a customer directly asks, the system must clearly state it is AI
    4. It depends on whether the company has disclosed AI use in its terms of service

13. What does channel consistency in AI disclosure mean for a company that uses AI in customer service emails?

    1. All emails must be sent from the same AI system
    2. Emails must contain AI disclosure in the same position as the website disclosure
    3. AI-generated emails should be disclosed with the same transparency as AI interactions on other channels
    4. Email disclosure is exempt because customers can unsubscribe

14. Why is 'legalese disclosure' insufficient even when it accurately describes AI use?

    1. Legal language is never permitted in consumer disclosures
    2. Customers cannot understand it and cannot act on it meaningfully
    3. It increases liability instead of reducing it
    4. Legal language makes the disclosure harder to find

15. A company's marketing team wants to label AI-powered personalization as 'smart recommendations.' What is the ethical concern?

    1. It implies AI is smarter than the customer
    2. It obscures AI involvement behind marketing language, preventing informed customer choice
    3. It could mislead competitors about the product's capabilities
    4. It may violate trademark law