Knowledge Check

The following question will help you check what you've learned.

1. Question 1 of 1A company notices that a customer-facing generative artificial intelligence (generative AI) chatbot is inadvertently outputting sensitive personally identifiable information from its training data. Which risk is being illustrated in this scenario?

   AJailbreaking: The chatbot is being "jailbroken" by prompts that trick it into bypassing its ethical constraints and content filters, permitting it to generate explicit or offensive content.BPrompt leaking: The chatbot's prompts and conversational context are being "leaked" to external sources, potentially exposing sensitive information or allowing unauthorized access.CExposure: The model's sensitive training data has been leaked or made publicly accessible, leading to data leaks and privacy violations.DHijacking: Malicious actors crafted prompts that have persuaded the model to generate harmful, unethical, or biased text outputs.
   Check answer

Navigate to the next lesson to explore resources related to prompt engineering.