Exam Bundle

A developer outside the release role cannot replace the artifact, while the expected digest loads successfully. This is evidence for a different corrective change: Enforce least-privilege artifact permissions, immutable versioning, and integrity checks before model deployment. That other change addresses: A trained model is saved in a broadly writable bucket where unrelated developers can replace weights without authorization. It does not validate the specific Beacon Health Network remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Authorized feature reads succeed and cross-team dataset access attempts are denied in the audit trail. This is evidence for a different corrective change: Restrict inference service access to its exact datasets and serving resources using narrowly scoped identities. That other change addresses: A production inference identity can list all organization datasets although it only requires one approved feature view. It does not validate the specific Beacon Health Network remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions. This is evidence for a different corrective change: Integrate appropriate Model Armor or supported safety and prompt-risk checks alongside least-privilege tool boundaries. That other change addresses: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. It does not validate the specific Beacon Health Network remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Pin reviewed container images and dependencies, verify provenance signatures or digests, and scan for known risks. This action could be suitable for a different engineering incident: A team deploys a model package from an unsigned third-party image with dependencies downloaded at runtime from uncontrolled mirrors. It does not directly correct the reported Brookfield Credit Union failure: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The discriminating acceptance check here is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. This action could be suitable for a different engineering incident: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not directly correct the reported Brookfield Credit Union failure: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The discriminating acceptance check here is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Enable an appropriate model-explainability method, document its limitations, and link explanations to model versions. This action could be suitable for a different engineering incident: A decision model provides probability scores but the governance report offers no reproducible local explanation for an adverse case. It does not directly correct the reported Brookfield Credit Union failure: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The discriminating acceptance check here is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

A retention test confirms expired personal prompts are deleted while required non-sensitive audit markers remain. This is evidence for a different corrective change: Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. That other change addresses: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not validate the specific Brookfield Credit Union remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

An authorized reviewer reproduces a local feature contribution for a frozen case without treating it as causal proof. This is evidence for a different corrective change: Enable an appropriate model-explainability method, document its limitations, and link explanations to model versions. That other change addresses: A decision model provides probability scores but the governance report offers no reproducible local explanation for an adverse case. It does not validate the specific Brookfield Credit Union remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

A tampered artifact is refused at release and the accepted image matches the approved supply-chain manifest. This is evidence for a different corrective change: Pin reviewed container images and dependencies, verify provenance signatures or digests, and scan for known risks. That other change addresses: A team deploys a model package from an unsigned third-party image with dependencies downloaded at runtime from uncontrolled mirrors. It does not validate the specific Brookfield Credit Union remediation for: A customer assistant includes complete account identifiers in LLM prompts although the response requires only purchase category. The necessary check instead is: Captured test requests contain only approved data fields and a secret-marker test never appears in model inputs.

Enforce least-privilege artifact permissions, immutable versioning, and integrity checks before model deployment. This action could be suitable for a different engineering incident: A trained model is saved in a broadly writable bucket where unrelated developers can replace weights without authorization. It does not directly correct the reported Atlas Equipment failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. This action could be suitable for a different engineering incident: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not directly correct the reported Atlas Equipment failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Measure group-sliced error and fairness metrics, investigate feature bias, and enforce an approved remediation process. This action could be suitable for a different engineering incident: A hiring predictor has good total accuracy but systematically rejects qualified applicants from one audited group. It does not directly correct the reported Atlas Equipment failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

A developer outside the release role cannot replace the artifact, while the expected digest loads successfully. This is evidence for a different corrective change: Enforce least-privilege artifact permissions, immutable versioning, and integrity checks before model deployment. That other change addresses: A trained model is saved in a broadly writable bucket where unrelated developers can replace weights without authorization. It does not validate the specific Atlas Equipment remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Held-out subgroup results meet the stated fairness guardrails without hiding deterioration in overall utility. This is evidence for a different corrective change: Measure group-sliced error and fairness metrics, investigate feature bias, and enforce an approved remediation process. That other change addresses: A hiring predictor has good total accuracy but systematically rejects qualified applicants from one audited group. It does not validate the specific Atlas Equipment remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

A retention test confirms expired personal prompts are deleted while required non-sensitive audit markers remain. This is evidence for a different corrective change: Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. That other change addresses: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not validate the specific Atlas Equipment remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Restrict inference service access to its exact datasets and serving resources using narrowly scoped identities. This action could be suitable for a different engineering incident: A production inference identity can list all organization datasets although it only requires one approved feature view. It does not directly correct the reported Crescent Utilities failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Document feature attributions as model-behavior explanations and require separate causal evidence for intervention claims. This action could be suitable for a different engineering incident: A business team presents feature attribution values from an opaque model as mathematical proof that changing the top-ranked attribute will change an outcome. It does not directly correct the reported Crescent Utilities failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Constrain each agent tool with separate scoped authorization and explicit approval for sensitive actions. This action could be suitable for a different engineering incident: An autonomous workflow agent can invoke billing exports using the same highly privileged identity it uses for ordinary summarization. It does not directly correct the reported Crescent Utilities failure: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The discriminating acceptance check here is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

A harmless summary succeeds but a disallowed export tool call is rejected and logged. This is evidence for a different corrective change: Constrain each agent tool with separate scoped authorization and explicit approval for sensitive actions. That other change addresses: An autonomous workflow agent can invoke billing exports using the same highly privileged identity it uses for ordinary summarization. It does not validate the specific Crescent Utilities remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Authorized feature reads succeed and cross-team dataset access attempts are denied in the audit trail. This is evidence for a different corrective change: Restrict inference service access to its exact datasets and serving resources using narrowly scoped identities. That other change addresses: A production inference identity can list all organization datasets although it only requires one approved feature view. It does not validate the specific Crescent Utilities remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

The model card labels explanatory associations accurately and rejects unsupported causal conclusions. This is evidence for a different corrective change: Document feature attributions as model-behavior explanations and require separate causal evidence for intervention claims. That other change addresses: A business team presents feature attribution values from an opaque model as mathematical proof that changing the top-ranked attribute will change an outcome. It does not validate the specific Crescent Utilities remediation for: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. The necessary check instead is: Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions.

Restrict inference service access to its exact datasets and serving resources using narrowly scoped identities. This action could be suitable for a different engineering incident: A production inference identity can list all organization datasets although it only requires one approved feature view. It does not directly correct the reported Horizon Grocery failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Integrate appropriate Model Armor or supported safety and prompt-risk checks alongside least-privilege tool boundaries. This action could be suitable for a different engineering incident: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. It does not directly correct the reported Horizon Grocery failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Measure group-sliced error and fairness metrics, investigate feature bias, and enforce an approved remediation process. This action could be suitable for a different engineering incident: A hiring predictor has good total accuracy but systematically rejects qualified applicants from one audited group. It does not directly correct the reported Horizon Grocery failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Authorized feature reads succeed and cross-team dataset access attempts are denied in the audit trail. This is evidence for a different corrective change: Restrict inference service access to its exact datasets and serving resources using narrowly scoped identities. That other change addresses: A production inference identity can list all organization datasets although it only requires one approved feature view. It does not validate the specific Horizon Grocery remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Held-out subgroup results meet the stated fairness guardrails without hiding deterioration in overall utility. This is evidence for a different corrective change: Measure group-sliced error and fairness metrics, investigate feature bias, and enforce an approved remediation process. That other change addresses: A hiring predictor has good total accuracy but systematically rejects qualified applicants from one audited group. It does not validate the specific Horizon Grocery remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Adversarial requests are blocked or constrained before protected tool invocation, with auditable policy decisions. This is evidence for a different corrective change: Integrate appropriate Model Armor or supported safety and prompt-risk checks alongside least-privilege tool boundaries. That other change addresses: An enterprise agent passes unsanitized untrusted user prompts to sensitive tools with no central policy enforcement. It does not validate the specific Horizon Grocery remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. This action could be suitable for a different engineering incident: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not directly correct the reported Silverline Insurance failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Validate trusted document provenance, restrict source updates, and record retrieval lineage before ingesting new context. This action could be suitable for a different engineering incident: A question-answering agent indexes mutable policy documents without content verification, allowing poisoned text to contaminate answers. It does not directly correct the reported Silverline Insurance failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

Enable an appropriate model-explainability method, document its limitations, and link explanations to model versions. This action could be suitable for a different engineering incident: A decision model provides probability scores but the governance report offers no reproducible local explanation for an adverse case. It does not directly correct the reported Silverline Insurance failure: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The discriminating acceptance check here is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

A forged policy document is rejected or quarantined and the agent cites only authorized versioned sources. This is evidence for a different corrective change: Validate trusted document provenance, restrict source updates, and record retrieval lineage before ingesting new context. That other change addresses: A question-answering agent indexes mutable policy documents without content verification, allowing poisoned text to contaminate answers. It does not validate the specific Silverline Insurance remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

A retention test confirms expired personal prompts are deleted while required non-sensitive audit markers remain. This is evidence for a different corrective change: Set data-minimizing storage and deletion policies across prompt, retrieval, logging, and model telemetry systems. That other change addresses: A legal-assistance assistant stores personal prompts indefinitely although contractual retention limits require rapid deletion. It does not validate the specific Silverline Insurance remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.

An authorized reviewer reproduces a local feature contribution for a frozen case without treating it as causal proof. This is evidence for a different corrective change: Enable an appropriate model-explainability method, document its limitations, and link explanations to model versions. That other change addresses: A decision model provides probability scores but the governance report offers no reproducible local explanation for an adverse case. It does not validate the specific Silverline Insurance remediation for: A generative assistant returns disallowed harmful instructions because a high overall helpfulness score is the only release metric. The necessary check instead is: A reviewed unsafe-prompt suite elicits the required refusals while permitted requests remain useful.