quiz bu lëkkale ak tegtal · Diggante Niveau
Cheapfakes vs Deepfakes Quiz
Distinguish simple edits and context changes from AI synthesis while verifying the underlying claim and respecting provenance limits.
Laaj 1 de 8
A genuine video is slowed down and recaptioned to change its meaning. Which description fits best?
Xoolaatal laaj bu nekk8 laaj
Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.
- A genuine video is slowed down and recaptioned to change its meaning. Which description fits best?
- A possible cheapfake or misleading edit
- A confirmed deepfake
- A verified accurate caption
- A C2PA credential
- A clip appears to show a public figure saying words absent from the full recording. What should be checked first?
- Call it a deepfake from appearance alone
- Find the full recording and compare it with the clip
- Assume the posted transcript is exact
- Share the clip before it disappears
- What difference is commonly meant by cheapfake versus deepfake?
- Cheapfakes are always false and deepfakes are always true
- Cheapfakes affect photos while deepfakes affect only audio
- Cheapfakes use relatively simple edits or context changes; deepfakes commonly involve AI synthesis or alteration
- Cheapfakes are legal and deepfakes are illegal
- A file contains valid C2PA Content Credentials. What can they support most directly?
- Proof that the caption accurately describes the event
- Proof that the publisher has no bias
- A guarantee that no misleading crop was shared
- Information about recorded provenance or edit history
- A repost has no Content Credentials. What can be concluded?
- Only that credentials are unavailable in this copy; absence does not prove manipulation
- The media is definitely synthetic
- The uploader intentionally removed them
- The scene is authentic
- An unusual facial expression appears in a compressed social clip. What is the responsible conclusion?
- The clip is certainly AI-generated
- The anomaly merits checking but does not alone establish a deepfake
- The speaker never made the statement
- Compression cannot affect appearance
- A real photograph is shared with the wrong event date. How should the issue be classified?
- A deepfake by definition
- A verified use of AI
- A context or caption problem, even if image pixels are genuine
- Proof the photograph itself is fabricated
- After comparing versions, you cannot determine whether AI was used. What should your report say?
- Declare it a deepfake to warn others
- Assume it is original because no detector flagged it
- Avoid explaining uncertainty
- Describe verified edits and state that the production method remains unknown