Identify Responsible AI and Data Protection Practices AB-730 Practice Questions (Page 2)

Questions 6–10

1. 6expert · hard

   A news organization uses a generative AI tool to draft articles. The editors are under pressure to publish quickly and have started publishing AI-generated articles without fact-checking. After a few weeks, they publish an article containing a fabricated quote from a public figure. What is the most likely cause of this incident?

   • AThe AI tool was trained on biased data.
   • BThe AI tool was not powerful enough to generate accurate content.
   • CThe editors over-relied on the AI tool without human oversight.
   • DThe AI tool was hacked by an external attacker.
   Reveal Solution Discussion Select an answer first
2. 7expert · hard

   A research team uses a generative AI tool to summarize scientific papers. The team plans to use the summaries in a grant proposal. To ensure the summaries are accurate, what is the most comprehensive verification approach?

   • AHave a domain expert review the summaries against the original papers.
   • BCompare the summaries with summaries generated by another AI tool.
   • CUse the summaries as-is because the AI tool is trained on scientific literature.
   • DAsk the AI tool to provide confidence scores for each summary.
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   Which of the following is the most accurate description of AI fabrication in the context of generative AI?

   • AThe model produces content that is factually incorrect or unsupported, presented with confidence.
   • BThe model refuses to answer a question due to safety filters.
   • CThe model repeats verbatim text from its training data.
   • DThe model intentionally lies to deceive the user.
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A software company uses a generative AI coding assistant. A developer pastes proprietary source code into the assistant to get suggestions for optimization. What is the primary risk of this action?

   • AThe AI assistant might suggest code that is inefficient.
   • BThe proprietary code could be exposed to the AI provider or other users.
   • CThe AI assistant will automatically file a patent for the code.
   • DThe AI assistant will delete the source code from the developer's machine.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A healthcare organization wants to use generative AI to draft patient education materials. They need to incorporate patient data to personalize the content. What is the most appropriate way to mitigate the risk of exposing sensitive patient information?

   • AEnter the data into the AI tool and rely on the tool's privacy policy.
   • BUse a public AI tool and remove patient names before entering data.
   • CUse an enterprise AI service with data residency and privacy commitments, and de-identify data before use.
   • DAvoid using any AI tool and write the materials manually.
   Reveal Solution Discussion Select an answer first