Snowflake SnowPro Specialty Gen AI GES-C01 Practice Test 42
Question 410
At a municipal utilities contractor, a Snowflake engineering team is undertaking a vendor compliance portal. An analysis app includes unmasked customer identifiers in an LLM prompt unnecessarily. The proposed fix is still hypothetical; document authors sometimes embed arbitrary instructions in descriptions, and audit requirements prohibit silent data leakage. Which TWO checks provide the most decisive evidence before approving a change?
- Inspect actual request payloads in a safe test
- Inspect trace fields and retention settings
- Compare token usage records to business job IDs
- Verify masking policy effects under the application role.
- Compare account region, cross-region parameters, and compliance policy
Explanation
Correct Answers
A:This observation directly tests the case mechanism: Warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. In the described setting, it helps determine whether minimize and mask identifying fields before any generative ai invocation. is justified. These author-intended interpretations require independent technical review before operational use.
D:This complementary observation tests a different part of the same hypothesis and should be read together with the first check before authorizing change. Expected success would be: the request includes only fields justified by the approved task. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:Checking this might help diagnose sensitive output logging, where a service traces entire user prompts and model outputs, accidentally retaining personal records for years. It is not one of the two most direct checks for the present case because the leading mechanism is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
C:Checking this might help diagnose token spend attribution, where a company sees rising ai credits but cannot identify which product line caused the spike. It is not one of the two most direct checks for the present case because the leading mechanism is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
E:Checking this might help diagnose data residency and cross-region, where a regulated tenant must keep processing inside a designated geography while a requested model is only available elsewhere. It is not one of the two most direct checks for the present case because the leading mechanism is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
Question 411
At a professional services consultancy, a Snowflake engineering team is undertaking a role-scoped analytics portal. An analysis app includes unmasked customer identifiers in an LLM prompt unnecessarily. Diagnosis is disputed because consultants work on multiple confidential clients; project billing must be attributable. Which underlying failure mechanism BEST explains the observed limitation?
- Warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values.
- Continued serving of the faulty configuration extends exposure.
- Automatic cross-region fallback can conflict with residency commitments.
- Overcollecting observability data creates a second sensitive repository.
Explanation
Correct Answer
A:This is the causal hypothesis consistent with the observed limitation: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Confirm it through both inspect actual request payloads in a safe test and verify masking policy effects under the application role. before changing the service. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:That mechanism could explain a different case about ai incident rollback, but does not most directly account for this specific symptom: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Here, the evidence to collect is inspect actual request payloads in a safe test These author-intended interpretations require independent technical review before operational use.
C:That mechanism could explain a different case about data residency and cross-region, but does not most directly account for this specific symptom: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Here, the evidence to collect is inspect actual request payloads in a safe test These author-intended interpretations require independent technical review before operational use.
D:That mechanism could explain a different case about sensitive output logging, but does not most directly account for this specific symptom: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Here, the evidence to collect is inspect actual request payloads in a safe test These author-intended interpretations require independent technical review before operational use.
Question 412
At an industrial maintenance company, a Snowflake engineering team is undertaking an asset fault knowledge base. An analysis app includes unmasked customer identifiers in an LLM prompt unnecessarily. The change was trialed on a small dataset; repair crews require concise results, while field logs contain abbreviations and exact part numbers. Which acceptance condition BEST demonstrates that the intervention worked?
- The request includes only fields justified by the approved task.
- Cost reports identify the source and trend of AI spend.
- Only the approved model identifiers are usable by the intended roles.
- Expired content is removed from all applicable stores under policy.
Explanation
Correct Answer
A:This is the observable acceptance condition that would show the intended fix addressed warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. It is stronger than activity counts because it tests the user-facing effect. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:This could demonstrate success for token spend attribution, but it would not establish resolution of the current issue: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Validate the request includes only fields justified by the approved task. instead. These author-intended interpretations require independent technical review before operational use.
C:This could demonstrate success for model allowlist policy, but it would not establish resolution of the current issue: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Validate the request includes only fields justified by the approved task. instead. These author-intended interpretations require independent technical review before operational use.
D:This could demonstrate success for data retention boundary, but it would not establish resolution of the current issue: an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Validate the request includes only fields justified by the approved task. instead. These author-intended interpretations require independent technical review before operational use.
Question 413
At a renewable-energy service provider, a Snowflake engineering team is undertaking an engineering knowledge assistant. An analysis app includes unmasked customer identifiers in an LLM prompt unnecessarily. An auditor will inspect the control design; the business needs verifiable source references, and technicians work in intermittent connectivity conditions. Which ongoing control MOST directly reduces the original risk?
- Integrate model allowlist changes into change control.
- Block outbound prompts containing sensitive markers.
- Enforce access controls and automatic retention expiry.
- Maintain emergency disable controls and evidence retention.
Explanation
Correct Answer
B:This retains a specific safeguard against recurrence of warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. It complements the immediate intervention, minimize and mask identifying fields before any generative ai invocation., and allows the team to verify the intended result. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This is a sensible safeguard for model allowlist policy, but it does not target the concrete exposure in this scenario. The current failure is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
C:This is a sensible safeguard for sensitive output logging, but it does not target the concrete exposure in this scenario. The current failure is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
D:This is a sensible safeguard for ai incident rollback, but it does not target the concrete exposure in this scenario. The current failure is warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values. These author-intended interpretations require independent technical review before operational use.
Question 414
At a non-profit social services network, a Snowflake engineering team is undertaking a service quality dashboard. An analysis app includes unmasked customer identifiers in an LLM prompt unnecessarily. Budget and operational risk are both material; some participant data is confidential; the organization serves populations with different access constraints. Which decision MOST appropriately balances the requested outcome with production constraints?
- Minimize and mask identifying fields before any generative AI invocation. Then block outbound prompts containing sensitive markers.
- Check the appropriate Snowflake Cortex database-role grants and assign only the function privileges it needs. Postpone the critical evidence checks until after a broad release.
- Treat retrieved content as untrusted evidence and test against instruction override attempts. Postpone the critical evidence checks until after a broad release.
- Disable the affected release, restore a reviewed prompt/model configuration, and audit impacted requests. Postpone the critical evidence checks until after a broad release.
Explanation
Correct Answer
A:This combines a corrective action with a bounded production safeguard, addressing an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. Prove the result by checking that the request includes only fields justified by the approved task. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:This proposal begins with an intervention for cortex role grant and postpones evidence that is essential for the current case. Before rollout, the team should minimize and mask identifying fields before any generative ai invocation. These author-intended interpretations require independent technical review before operational use.
C:This proposal begins with an intervention for prompt injection threat modeling and postpones evidence that is essential for the current case. Before rollout, the team should minimize and mask identifying fields before any generative ai invocation. These author-intended interpretations require independent technical review before operational use.
D:This proposal begins with an intervention for ai incident rollback and postpones evidence that is essential for the current case. Before rollout, the team should minimize and mask identifying fields before any generative ai invocation. These author-intended interpretations require independent technical review before operational use.
Question 415
At a university research office, a Snowflake engineering team is undertaking a restricted data discovery project. A shared search service indexes HR contracts that contain protected employee details. Before implementation, staff rotate between roles during the academic year; research access differs by grant. Which first architectural or operating step MOST directly addresses this case?
- Segment evaluation by language and user cohort and investigate disparities before launch.
- Verify factual assertions against actual retrieved records before distributing the report.
- Restrict source material and retrieval results to permitted audiences before deployment.
- Minimize and mask identifying fields before any generative AI invocation.
Explanation
Correct Answer
C:The decision addresses the central constraint: a shared search service indexes hr contracts that contain protected employee details. The most direct next step is to restrict source material and retrieval results to permitted audiences before deployment. Before changing production, review both: run access-negative searches for hr files and verify role-specific filters.. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This can be appropriate when a chatbot scores highly overall but fails much more often on one product language. It addresses ai evaluation bias, whereas the present limitation is better explained by indexing protected data can expose snippets if retrieval authorization is weak. The target outcome here is that unauthorized roles cannot retrieve confidential excerpts or identifiers. These author-intended interpretations require independent technical review before operational use.
B:This can be appropriate when a generated audit summary cites a warehouse table that was never queried. It addresses model output hallucination, whereas the present limitation is better explained by indexing protected data can expose snippets if retrieval authorization is weak. The target outcome here is that unauthorized roles cannot retrieve confidential excerpts or identifiers. These author-intended interpretations require independent technical review before operational use.
D:This can be appropriate when an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. It addresses data masking before prompts, whereas the present limitation is better explained by indexing protected data can expose snippets if retrieval authorization is weak. The target outcome here is that unauthorized roles cannot retrieve confidential excerpts or identifiers. These author-intended interpretations require independent technical review before operational use.
Question 416
At a community-bank technology team, a Snowflake engineering team is undertaking an internal policy assistant pilot. A shared search service indexes HR contracts that contain protected employee details. The proposed fix is still hypothetical; security rejects any unapproved cross-border data processing, and the pilot must serve two departments with different permissions. Which TWO checks provide the most decisive evidence before approving a change?
- Inspect trace fields and retention settings
- Run access-negative searches for HR files
- Compare token usage records to business job IDs
- Verify role-specific filters.
- Inspect actual request payloads in a safe test
Explanation
Correct Answers
B:This observation directly tests the case mechanism: Indexing protected data can expose snippets if retrieval authorization is weak. In the described setting, it helps determine whether restrict source material and retrieval results to permitted audiences before deployment. is justified. These author-intended interpretations require independent technical review before operational use.
D:This complementary observation tests a different part of the same hypothesis and should be read together with the first check before authorizing change. Expected success would be: unauthorized roles cannot retrieve confidential excerpts or identifiers. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:Checking this might help diagnose sensitive output logging, where a service traces entire user prompts and model outputs, accidentally retaining personal records for years. It is not one of the two most direct checks for the present case because the leading mechanism is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
C:Checking this might help diagnose token spend attribution, where a company sees rising ai credits but cannot identify which product line caused the spike. It is not one of the two most direct checks for the present case because the leading mechanism is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
E:Checking this might help diagnose data masking before prompts, where an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. It is not one of the two most direct checks for the present case because the leading mechanism is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
Question 417
At an e-commerce marketplace, a Snowflake engineering team is undertaking a returns reason classification effort. A shared search service indexes HR contracts that contain protected employee details. Diagnosis is disputed because some tickets mention several products; billing and privacy queues need distinct ownership. Which underlying failure mechanism BEST explains the observed limitation?
- Continued serving of the faulty configuration extends exposure.
- Aggregate performance hides concentrated failures for underrepresented users.
- Indexing protected data can expose snippets if retrieval authorization is weak.
- Lowest per-token price may increase downstream errors and remediation spend.
Explanation
Correct Answer
C:This is the causal hypothesis consistent with the observed limitation: a shared search service indexes hr contracts that contain protected employee details. Confirm it through both run access-negative searches for hr files and verify role-specific filters. before changing the service. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:That mechanism could explain a different case about ai incident rollback, but does not most directly account for this specific symptom: a shared search service indexes hr contracts that contain protected employee details. Here, the evidence to collect is run access-negative searches for hr files These author-intended interpretations require independent technical review before operational use.
B:That mechanism could explain a different case about ai evaluation bias, but does not most directly account for this specific symptom: a shared search service indexes hr contracts that contain protected employee details. Here, the evidence to collect is run access-negative searches for hr files These author-intended interpretations require independent technical review before operational use.
D:That mechanism could explain a different case about inference cost versus quality, but does not most directly account for this specific symptom: a shared search service indexes hr contracts that contain protected employee details. Here, the evidence to collect is run access-negative searches for hr files These author-intended interpretations require independent technical review before operational use.
Question 418
At an insurance claims processor, a Snowflake engineering team is undertaking a form intake modernization. A shared search service indexes HR contracts that contain protected employee details. The change was trialed on a small dataset; the new system cannot automatically deny claims, while manual reviewers handle ambiguous scanned content. Which acceptance condition BEST demonstrates that the intervention worked?
- Required diagnostics remain available without unnecessary personal data.
- The assistant follows trusted policy while extracting only relevant facts.
- Unauthorized roles cannot retrieve confidential excerpts or identifiers.
- Only the approved model identifiers are usable by the intended roles.
Explanation
Correct Answer
C:This is the observable acceptance condition that would show the intended fix addressed indexing protected data can expose snippets if retrieval authorization is weak. It is stronger than activity counts because it tests the user-facing effect. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This could demonstrate success for sensitive output logging, but it would not establish resolution of the current issue: a shared search service indexes hr contracts that contain protected employee details. Validate unauthorized roles cannot retrieve confidential excerpts or identifiers. instead. These author-intended interpretations require independent technical review before operational use.
B:This could demonstrate success for prompt injection threat modeling, but it would not establish resolution of the current issue: a shared search service indexes hr contracts that contain protected employee details. Validate unauthorized roles cannot retrieve confidential excerpts or identifiers. instead. These author-intended interpretations require independent technical review before operational use.
D:This could demonstrate success for model allowlist policy, but it would not establish resolution of the current issue: a shared search service indexes hr contracts that contain protected employee details. Validate unauthorized roles cannot retrieve confidential excerpts or identifiers. instead. These author-intended interpretations require independent technical review before operational use.
Question 419
At a consumer travel platform, a Snowflake engineering team is undertaking a booking assistant redesign. A shared search service indexes HR contracts that contain protected employee details. An auditor will inspect the control design; a bad response can trigger a financial adjustment, and peak demand is highly seasonal. Which ongoing control MOST directly reduces the original risk?
- Version and monitor regional restrictions.
- Integrate model allowlist changes into change control.
- Couple index membership and search filters to document ACLs.
- Make fairness testing part of model release gates.
Explanation
Correct Answer
C:This retains a specific safeguard against recurrence of indexing protected data can expose snippets if retrieval authorization is weak. It complements the immediate intervention, restrict source material and retrieval results to permitted audiences before deployment., and allows the team to verify the intended result. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This is a sensible safeguard for data residency and cross-region, but it does not target the concrete exposure in this scenario. The current failure is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
B:This is a sensible safeguard for model allowlist policy, but it does not target the concrete exposure in this scenario. The current failure is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
D:This is a sensible safeguard for ai evaluation bias, but it does not target the concrete exposure in this scenario. The current failure is indexing protected data can expose snippets if retrieval authorization is weak. These author-intended interpretations require independent technical review before operational use.
Question 420
At a media-subscription company, a Snowflake engineering team is undertaking a finance analytics chatbot. A shared search service indexes HR contracts that contain protected employee details. Budget and operational risk are both material; financial metrics have overlapping business synonyms; board reporting requires certified definitions. Which decision MOST appropriately balances the requested outcome with production constraints?
- Couple index membership and search filters to document ACLs.
- Restrict source material and retrieval results to permitted audiences before deployment.
- Make fairness testing part of model release gates.
- Require an evidence manifest for material assertions.
- Block outbound prompts containing sensitive markers.
Explanation
Correct Answers
A:This establishes the ongoing safety boundary after the immediate correction. It provides protection while the team checks that unauthorized roles cannot retrieve confidential excerpts or identifiers. These author-intended interpretations require independent technical review before operational use.
B:This implements the response to the stated failure: Indexing protected data can expose snippets if retrieval authorization is weak. It must be validated using run access-negative searches for hr files These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
C:This guardrail primarily addresses ai evaluation bias and is not the companion intervention the case most urgently requires. The stated issue is a shared search service indexes hr contracts that contain protected employee details. These author-intended interpretations require independent technical review before operational use.
D:This guardrail primarily addresses model output hallucination and is not the companion intervention the case most urgently requires. The stated issue is a shared search service indexes hr contracts that contain protected employee details. These author-intended interpretations require independent technical review before operational use.
E:This guardrail primarily addresses data masking before prompts and is not the companion intervention the case most urgently requires. The stated issue is a shared search service indexes hr contracts that contain protected employee details. These author-intended interpretations require independent technical review before operational use.