Creating Malicious Software with AI GOAA Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   Which of the following is a common technique used to detect deepfake videos?

   • AAnalyzing blinking patterns and facial inconsistencies.
   • BVerifying the video's frame rate.
   • CChecking the file size of the video.
   • DLooking at the video's resolution.
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   Which technique is commonly used in deepfake creation to swap one person's face onto another's body in a video?

   • AUsing an autoencoder to compress and reconstruct facial features, then blending the reconstructed face onto the target video.
   • BManually editing each frame with a photo editor.
   • CApplying a simple filter to change the lighting and color of the video.
   • DUsing optical character recognition to overlay text on the video.
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A penetration tester is hired to assess a client's defenses against AI-generated phishing emails. The tester plans to use a language model to craft convincing phishing messages. Which action is most ethically and legally appropriate?

   • AUse the AI to generate emails but manually review them to ensure they do not contain any malicious attachments.
   • BObtain written authorization from the client that explicitly permits the use of AI-generated content in the phishing simulation.
   • CDeploy the AI-generated emails to a small test group outside the client's organization to avoid impacting internal systems.
   • DUse the AI-generated emails only against the client's own employees, without informing them, to ensure realism.
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   What is a known limitation of deepfake detection methods?

   • AThey may not generalize to new deepfake generation techniques.
   • BThey always require high-resolution video to work.
   • CThey are computationally inexpensive and run in real time on any device.
   • DThey can only detect deepfakes of faces, not other objects.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A video forensics analyst is reviewing a video that is suspected to be a deepfake. The analyst notices that the subject's head movements are smooth but the facial expressions do not match the spoken words. Which detection technique would be most effective in confirming the manipulation?

   • AUse a GAN discriminator to classify the video as real or fake.
   • BExamine the video for compression artifacts that are typical of deepfake generation.
   • CAnalyze the audio-visual synchronization to detect mismatches between lip movements and speech.
   • DCheck the video's metadata for editing software signatures.
   Reveal Solution Discussion Select an answer first