AI Stock-Photo Disclosure: Marketplace Provenance Standards

End-of-lesson check

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

1. Which of the following is typically included in a C2PA-compliant manifest entry for an AI-generated image?

   1. The complete training dataset used to train the generation model
   2. The IP addresses of all users who viewed the image
   3. The seller's payment information and pricing history
   4. The model name, version, and a summary of the prompt used

2. According to best practices for AI-generated stock photos, what should a seller checklist include before listing an asset?

   1. A list of all buyers who have expressed interest in similar images
   2. The email addresses of the original model developers
   3. The exact GPU hardware used during generation
   4. The model name, a prompt summary, and disclosure of any post-generation edits

3. Which task can AI reliably perform in the context of stock-photo marketplace disclosure?

   1. License the model's training data rights to downstream buyers
   2. Verify that the seller actually used the disclosed model to create the asset
   3. Determine appropriate indemnity caps for the platform
   4. Generate C2PA-compliant manifest entries for each upload

4. What is a key limitation of AI in managing stock-photo marketplace disclosures?

   1. AI cannot identify which images were AI-generated
   2. AI cannot generate manifest entries in a standardized format
   3. AI cannot determine indemnity caps appropriate to the platform
   4. AI cannot redact PII from prompt summaries

5. A seller properly discloses the model used to generate an image. What additional right must buyers obtain separately?

   1. A guarantee that no similar images will be generated in the future
   2. The right to modify the image without attribution
   3. The right to resell the image on competing marketplaces
   4. A license to the model's training data rights

6. When a seller's claim about an image conflicts with embedded EXIF data, what should a marketplace system flag?

   1. The image should be automatically deleted
   2. The conflict should be flagged for review
   3. The seller should be paid immediately
   4. The buyer should be blocked from downloading

7. What does the term 'provenance' refer to in the context of AI-generated stock photos?

   1. The market price history of similar images
   2. The artistic style and aesthetic quality of the image
   3. The chronological record of an asset's origin, creation, and ownership history
   4. The geographic location where the image was uploaded

8. What is a 'human-author claim' in the context of AI-generated asset disclosure?

   1. A statement that no AI tools were used in post-processing
   2. A claim that substantial human creative input justifies treating the AI output as human-authored for copyright purposes
   3. A guarantee that the image contains no AI-generated elements
   4. A declaration that a human photographer took the image

9. Why should prompt summaries in disclosure metadata be redacted of PII?

   1. Redaction is required by the model developers' terms of service
   2. PII makes the manifest non-C2PA-compliant
   3. PII in prompts could reveal the seller's identity to competitors
   4. To protect personally identifiable information from being publicly exposed in the metadata

10. What should be disclosed about 'post-edits' when listing an AI-generated image on a stock marketplace?

    1. Only edits that were performed by professional designers
    2. Only edits that changed the subject matter of the image
    3. The specific software tools used for editing
    4. Any human modifications made after the AI generation process, such as color adjustments or cropping

11. In the context of AI-generated stock photos, what does 'indemnity' typically cover?

    1. Legal liability and damages arising from the use of the asset, such as copyright claims
    2. The computational cost of generating the image
    3. The fee charged by the marketplace for listing the image
    4. The cost of hosting the image on marketplace servers

12. What is required for a stock-photo marketplace to achieve 'C2PA compliance' for AI-generated assets?

    1. Storing all original AI prompts on a blockchain
    2. Paying a licensing fee to the C2PA organization
    3. Using only human photographers for all images
    4. Implementing technical standards for content provenance and authenticity metadata

13. A seller uploads an AI-generated image but claims it was entirely created by a human artist. The EXIF data shows AI generation parameters. What is the appropriate system response?

    1. Add a generic 'AI-assisted' label to the image
    2. Automatically reject the listing without explanation
    3. Approve the listing since the seller is the content owner
    4. Flag the discrepancy between the seller claim and EXIF data for review

14. What is meant by 'inherited risk' in the context of buying AI-generated stock photos?

    1. The computational cost of running the AI model
    2. Potential legal liabilities that transfer with the image, such as undisclosed training data infringements
    3. The risk that similar images will be generated by other users
    4. The risk that the image will become obsolete quickly

15. Which of the following statements accurately reflects what AI tools can do for stock-photo marketplace disclosure?

    1. AI can generate C2PA-compliant manifest entries automatically
    2. AI can verify that sellers are honestly reporting the models they used
    3. AI can guarantee that no copyright claims will arise from use
    4. AI can determine the fair market value of the asset