Gids-gekoppelde quiz · Middelmatig Niveau

AI-inhoudslabels op sociale platformsquiz

Test hoe platforms AI-inhoud detecteren via C2PA- en IPTC-metadata en watermerken, wat makers moeten vrijgeven en waarom labels vaak ontbreken.

Vraag 1 van 8

Waarom heeft Meta de naam 'Made with AI' gewijzigd in 'AI info'?

Bekijk elke vraag8 vragen

Antwoorden en uitleg verschijnen terwijl je de quiz doet.

  1. Waarom heeft Meta de naam 'Made with AI' gewijzigd in 'AI info'?
    • A court ordered the change
    • The label was too small to read
    • Photographers complained that minor AI-assisted edits triggered the label on real photos
    • Meta stopped labeling AI content entirely
  2. Welke inhoud moeten makers van YouTube openbaar maken?
    • Realistic altered or synthetic content, such as a real person appearing to say something they did not
    • Any video that uses a beauty filter
    • Any video whose script was drafted with AI
    • Only clearly unrealistic animated fantasy content
  3. Op welke signalen vertrouwen platforms vooral om AI-inhoud automatisch te detecteren?
    • The number of likes a post receives
    • Metadata such as C2PA and IPTC fields, plus invisible watermarks
    • The creator's follower count
    • The time of day the post was uploaded
  4. Waarom krijgt een screenshot van een AI-afbeelding geen automatisch label?
    • Platforms never label images
    • Screenshots are automatically marked as authentic
    • Screenshots are always lower resolution than the platform accepts
    • Taking a screenshot creates a new file without the original metadata
  5. Welke IPTC digitalSourceType-waarde markeert volledig door AI gegenereerde inhoud?
    • trainedAlgorithmicMedia
    • compositeWithTrainedAlgorithmicMedia
    • digitalCapture
    • originalPhotograph
  6. What does the absence of an AI label on a post tell you?
    • The content is definitely real
    • The platform verified it with a human reviewer
    • It is not proof that the content is real
    • The content was made before 2020
  7. How is a C2PA manifest tied to the file it describes?
    • By the file name
    • By the upload date
    • By a visible logo in the corner
    • By a cryptographic hash and signature, so tampering is detectable
  8. What does Article 50 of the EU AI Act cover, according to the guide?
    • A ban on all AI-generated images
    • Transparency obligations such as marking synthetic outputs and disclosing deepfakes, set to apply from August 2026
    • Mandatory payment for every AI image posted
    • Rules only for AI chatbots in banking