Identify Responsible AI and Data Protection Practices AB-730 Practice Questions (Page 8)

Questions 36–40

1. 36expert · hard

   A government agency wants to use a generative AI tool to assist in drafting public communications. The tool is a cloud service. The agency handles sensitive citizen data and is subject to strict data protection regulations. The agency's IT team is concerned about prompt injection attacks that could leak sensitive data. What is the most comprehensive approach?

   • AUse the cloud service but only for non-sensitive communications.
   • BUse the cloud service but restrict the AI's access to only non-sensitive data.
   • CUse the cloud service and rely on the provider's security measures.
   • DDeploy the AI tool on-premises and implement input/output filtering to prevent prompt injection.
   Reveal Solution Discussion Select an answer first
2. 37application · medium

   A marketing analyst uses a generative AI tool to draft a blog post about a new product. The draft includes a statistic about market share that the analyst has never seen before. The analyst wants to use the draft but is concerned about accuracy. What should the analyst do first?

   • AAsk the AI tool to generate a citation for the statistic and then publish.
   • BSearch for the statistic in reputable industry reports before including it.
   • CRemove the statistic and replace it with a similar one the analyst remembers from a conference.
   • DPublish the draft as-is because the AI tool is generally reliable.
   Reveal Solution Discussion Select an answer first
4. 38foundation · easy

   Which of the following is a recommended strategy to mitigate sensitive data risks when using generative AI?

   • AImplement data loss prevention (DLP) policies that block sensitive data from being entered into AI tools.
   • BEncourage employees to use personal accounts for AI tools to avoid corporate tracking.
   • CStore all AI outputs in a shared public folder for transparency.
   • DAllow all employees to use any AI tool without restrictions.
   Reveal Solution Discussion Select an answer first
5. 39application · medium

   A customer support team uses an AI tool to generate responses to customer emails. The team has stopped reviewing the AI-generated responses before sending them. What is the most important risk of this practice?

   • AThe AI will send emails to the wrong recipients.
   • BThe AI will increase the team's workload because it needs constant maintenance.
   • CThe AI may generate responses that contain inaccurate or inappropriate information.
   • DThe AI will eventually run out of training data and stop working.
   Reveal Solution Discussion Select an answer first
6. 40expert · hard

   A pharmaceutical company uses a generative AI tool to summarize clinical trial results. The AI generates a summary that claims a new drug is highly effective, but the summary does not mention any side effects. The company plans to submit the summary to a regulatory agency. What is the most appropriate action?

   • ASubmit the summary to the regulatory agency and let them verify the data.
   • BAsk the AI tool to generate a more detailed summary that includes side effects.
   • CHave a medical expert review the summary against the full trial data before submission.
   • DSubmit the summary as-is because the AI tool is known for accuracy.
   Reveal Solution Discussion Select an answer first