Creating Malicious Software with AI GOAA Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A cybersecurity consultant is asked by a client to create a deepfake of a competitor's CEO to test the client's social engineering defenses. What should the consultant do?

   • ACreate the deepfake but ensure it is clearly labeled as a fake in all materials.
   • BCreate the deepfake using a consenting actor's face instead of the competitor's CEO.
   • CCreate the deepfake but only use it internally within the client's organization.
   • DRefuse the request because creating a deepfake of a real person without their consent is unethical and likely illegal.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A fact-checking organization is investigating a video that allegedly shows a politician making a controversial statement. The video has been shared widely. Which detection technique would be most useful to determine if the video is a deepfake?

   • ASearch for the video on social media to see if it has been debunked before.
   • BAsk the politician's office to confirm whether the statement was made.
   • CCheck the video's file size and resolution to see if it matches typical deepfake outputs.
   • DAnalyze the video for inconsistencies in blinking, head movements, and lighting.
   Reveal Solution Discussion Select an answer first
4. 3expert · hard

   A red team is tasked with testing an organization's defenses against AI-generated malware. They have access to a generative model that can produce new malware variants. The organization's defenses include both signature-based antivirus and an AI-based behavioral detector. Which strategy would be most effective for the red team to evaluate the AI-based detector?

   • ATest the AI detector with benign files that are slightly modified to appear malicious.
   • BFocus on creating variants that are completely novel and have no resemblance to known malware.
   • CGenerate variants that are functionally equivalent to known malware but have different code structures to evade signature detection, then test them against the AI detector.
   • DUse the generative model to create malware that is specifically designed to be undetectable by any security tool.
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   A digital forensics investigator is examining a video that may be a deepfake. The video shows a CEO making a statement, but the lighting on the face appears inconsistent with the background. Which deepfake creation technique is most likely to produce such an artifact?

   • AA voice cloning model that synthesizes the CEO's speech and is then dubbed over the original video.
   • BAn autoencoder that reconstructs the entire frame from a compressed latent representation.
   • CA simple video editing tool that overlays a static image of the CEO's face onto the original footage.
   • DA GAN-based face-swap model that generates the face region independently from the original lighting conditions.
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A red-team engineer is tasked with generating a polymorphic dropper that evades signature-based antivirus while maintaining functionality. The team has access to a generative model fine-tuned on benign PE files. Which approach best leverages the model while minimizing the risk of producing non-functional code?

   • AUse the model to generate a new dropper for each target machine, skipping sandbox testing to save time.
   • BUse the model to generate entirely new dropper binaries from scratch, then test each one in a sandbox.
   • CUse the model as an instruction-tuned code generator to produce a dropper, then manually review and compile it.
   • DSeed the model with a known-good dropper and prompt it to apply semantic-preserving transformations, then verify functionality in a sandbox.
   Reveal Solution Discussion Select an answer first