SnowPro Specialty: AI Governance & Security free question 6 / 10 · Domain 3: AI Guardrails and Content Safety
A reviewer wants to see the effect of Cortex Guard on a set of prompts in Snowsight before anyone writes SQL. Where does the Cortex Playground documentation put the switch?
- In the settings panel as the Enable Cortex Guard checkbox next to the other model settings
- In the Required privileges section as a grant to the role that opens the playground
- In the Connect your data dialog as an option on the column that holds the text
- In the warehouse picker as a setting that applies to the warehouse that runs the prompt
Show the answer and explanation
Answer: A
Why this is correct
The Cortex Playground page lists Enable Cortex Guard among the settings that can be changed with Change settings. It sits next to temperature, top_p, and max_tokens, and it filters out potentially inappropriate or unsafe responses from the model.
Why the other options are wrong
"In the Required privileges section as a grant to the role that opens the playground" … The Required privileges section names the CORTEX_USER database role for calling the LLM functions. It does not hold a Cortex Guard switch.
"In the Connect your data dialog as an option on the column that holds the text" … The data dialog connects a table column to the playground. The page puts the safeguard in the settings of the model.
"In the warehouse picker as a setting that applies to the warehouse that runs the prompt" … The warehouse runs the SQL command that calls COMPLETE. The page does not describe a warehouse setting for Cortex Guard.
A try-out switch for the same feature
Enable Cortex Guard is among the settings that can be changed with Change settings in the Cortex Playground.