Question 380
A phishing analyzer relies on a domain-age heuristic that attackers now evade using established domains. Which intelligence practice would solve this problem most directly?
- Adversary-adaptation feature review
- Contextual false-positive tuning
- Precision-recall trade-off assessment
- Coverage-stratified model evaluation
Explanation
Correct Answer
A:Feature drift should trigger reevaluation of behavior and context signals rather than a permanent dependence on domain age.
Incorrect Answers
B:Allowlisted software identity and normal destination patterns can reduce benign matches without disabling all periodicity detection; although useful for preserving signal while excluding expected polling, it is not an adequate solution for responding to attackers' use of aged infrastructure.
C:Threshold changes alter false positives and false negatives; select operating points based on consequence and review capacity; the distinction is between balancing detection sensitivity with analyst workload and responding to attackers' use of aged infrastructure; the latter governs this decision.
D:Metrics by language and geography expose weak subpopulations that an aggregate accuracy figure can conceal; the specified requirement concerns responding to attackers' use of aged infrastructure, while this option supports locating language-specific detection gaps.
Question 381
A source gives a domain confidence score of 95 with no indication of scale or calibration. Which procedure would most effectively close the stated gap?
- Feed collateral-impact review
- Confidence-scale semantics review
- Novelty and overlap scoring
- Membership-and-handling review
Explanation
Correct Answer
B:Numeric confidence is meaningful only with stated definition, calibration and context; scores from different providers are not automatically comparable.
Incorrect Answers
A:Shared hosting and recycled addresses increase false-positive cost; feed adoption should consider reversibility and business impact; the immediate priority is interpreting a vendor-specific numeric score; evaluating the safety of automated IP enforcement is a distinct activity.
C:Measuring unique, useful records after normalization reveals marginal value beyond simple published volume; this is useful for quantifying whether a new source adds coverage, but it cannot replace the practice needed for interpreting a vendor-specific numeric score.
D:A restricted sharing program can yield unique indicators but requires adherence to contribution, recipient and confidentiality terms; those results can inform judging suitability of access-controlled community intelligence without satisfying the need for interpreting a vendor-specific numeric score.
Question 382
A malware team wants to scan thousands of unexecuted samples for a distinctive combination of byte strings and file conditions. Which procedure would most effectively close the stated gap?
- STIX schema validation tooling
- YARA-based sample classification
- ATT&CK Navigator layer comparison
- Email header analysis tooling
Explanation
Correct Answer
B:YARA rules test declared file-content patterns and logical conditions, allowing scalable triage while requiring validation against benign lookalikes.
Incorrect Answers
A:Typed validation catches malformed object identifiers, missing properties and invalid relationships before they contaminate downstream graphs; it does not assess threat truth; here the criterion is classifying stored binaries using static signature patterns, not simply rejecting improperly structured shared intelligence.
C:Technique layers can display coverage or priorities on a matrix, making gaps visible without asserting a technique is detected simply because it is colored; that supports visualizing behavior coverage with explicit evidence, whereas the task requires classifying stored binaries using static signature patterns.
D:Header parsers reveal routing and SPF/DKIM/DMARC results, which complement but do not replace content analysis; although useful for investigating message provenance and authentication, it is not an adequate solution for classifying stored binaries using static signature patterns.
Question 383
An intelligence pipeline forwards cloud logs into a shared store without tagging account IDs. How should the threat intelligence team proceed?
- Tenant-safe cloud intelligence sanitization
- Workload-ID and orchestration metadata enrichment
- Account- and tenant-identity normalization
- Cross-account trust-policy enrichment
Explanation
Correct Answer
C:Preserving source account and tenant keys avoids false correlations when resource names repeat across environments.
Incorrect Answers
A:Replace unnecessary identifiers and enforce policy before shared export, while preserving private linkage for responders; the case calls for preventing cross-account entity confusion, which does not follow from sharing cloud trends without exposing tenant-specific data alone.
B:Stable deployment, namespace and service-account labels provide context when short-lived container IDs disappear; it can assist with correlating activity from ephemeral cloud workloads; the case still requires preventing cross-account entity confusion.
D:Identity and resource policy context helps determine why a principal from another account could assume or use access; this works toward explaining trust relationships behind API calls; the described problem instead calls for preventing cross-account entity confusion.
Question 384
A TIP ingests intelligence with no source identifiers and cannot explain why a domain was flagged. What should be selected to fulfill this intelligence need?
- Vocabulary mapping and controlled categories
- Provenance metadata preservation
- Entity alias and synonym indexing
- Indicator syntax and context validation
Explanation
Correct Answer
B:Source attribution and acquisition time must survive transformation so conclusions can be traced and withdrawn when needed.
Incorrect Answers
A:Documented mappings reconcile synonyms and differing source schemas without inventing unsupported equivalence; that activity pertains to comparing classifications from incompatible publishers, rather than the requested retaining origin throughout processing.
C:Maintained alias mappings improve retrieval while preserving uncertainties in identity rather than collapsing all similar labels; here the criterion is retaining origin throughout processing, not simply finding reports that use alternate names for an entity.
D:Field validation catches invalid formats and impossible contexts before noisy or unsafe records reach enforcement; this would help with rejecting malformed observables during processing, but would not by itself address retaining origin throughout processing.
Question 385
An analyst wants a representative view of a huge dataset but full export would overwhelm available compute. What should be selected to fulfill this intelligence need?
- Selective field and event filtering
- Stratified sampling for exploratory review
- Idempotent checkpointed ingestion
- Atomic ingest and completion markers
- Immutable raw storage plus derived index
Explanation
Correct Answer
B:Sampling across meaningful classes or time bands can expose patterns without pretending the sample is the complete population.
Incorrect Answers
A:Collecting required fields and relevant event classes first reduces storage pressure without discarding evidence needed by PIRs; its purpose is controlling volume while retaining decision-relevant data, while the present requirement concerns examining large data while preserving subgroup representation.
C:Stable record identifiers and committed progress markers allow safe replay without multiplying previously accepted records; the case calls for examining large data while preserving subgroup representation, which does not follow from resuming interrupted bulk jobs accurately alone.
D:Publishing a dataset only after full receipt and validation prevents downstream parsers from consuming truncated files; it can assist with ensuring batch consumers see only complete acquisitions; the case still requires examining large data while preserving subgroup representation.
E:Separating an unchanged acquisition copy from processed search data preserves provenance while enabling efficient analysis; the immediate priority is examining large data while preserving subgroup representation; retaining originals alongside normalized data is a distinct activity.
Question 386
A dataset includes malformed hashes and non-routable test IPs posing as high-confidence indicators. What should be selected to fulfill this intelligence need?
- Context-sensitive indicator extraction
- IP address representation normalization
- Typed schema reconciliation
- Indicator syntax and context validation
Explanation
Correct Answer
D:Field validation catches invalid formats and impossible contexts before noisy or unsafe records reach enforcement.
Incorrect Answers
A:Section context, negation and validation can prevent automatic promotion of illustrative or harmless observables; its main application is avoiding overcollection from narrative source material, not rejecting malformed observables during processing.
B:An explicit parsing and canonicalization policy prevents duplicate records or missed matches across supported address representations; it addresses reconciling equivalent network address formats; the situation instead calls for rejecting malformed observables during processing.
C:Keep metric type, units and scale alongside values so unrelated measures are not compared as if interchangeable; that activity pertains to preventing invalid cross-source metric comparisons, rather than the requested rejecting malformed observables during processing.
Question 387
An extractor reads a vendor PDF and mistakes a footnote's benign IP for a malicious IOC. Which action or assessment best matches the need?
- Context-sensitive indicator extraction
- Safe IOC defanging reversal
- Indicator syntax and context validation
- Shared-infrastructure context enrichment
Explanation
Correct Answer
A:Section context, negation and validation can prevent automatic promotion of illustrative or harmless observables.
Incorrect Answers
B:Controlled conversion restores parsable URLs for analysis while ensuring displayed or clicked links are not accidentally activated; this provides evidence relevant to parsing deliberately deactivated indicators securely; the remaining need is avoiding overcollection from narrative source material.
C:Field validation catches invalid formats and impossible contexts before noisy or unsafe records reach enforcement; its purpose is rejecting malformed observables during processing, while the present requirement concerns avoiding overcollection from narrative source material.
D:Ownership and co-tenancy context help avoid interpreting address co-location as evidence of a single campaign; this supports interpreting address associations cautiously, not the technical control required for avoiding overcollection from narrative source material.
Question 388
A new team member cannot tell which knowledge-base entries are verified and which are informal guesses. Which action or assessment best matches the need?
- Runbook false-positive examples
- Knowledge article validation status
- Query-and-evidence knowledge capture
- Reusable analyst knowledge article
Explanation
Correct Answer
B:Review dates, confidence and source links distinguish tested procedures from tentative field notes. It directly addresses separating approved instructions from provisional ideas.
Incorrect Answers
A:Known-good patterns, counterexamples and escalation criteria make searches operationally interpretable instead of merely executable; those results can inform helping analysts discriminate benign from suspicious results without satisfying the need for separating approved instructions from provisional ideas.
C:Saving validated searches, schema assumptions and example observations enables defenders to reproduce the finding; its main application is turning an insight into reusable detection work, not separating approved instructions from provisional ideas.
D:Documented methods, examples, source limitations and validation steps make a successful research approach transferable; it addresses capturing investigative know-how for colleagues; the situation instead calls for separating approved instructions from provisional ideas.
Question 389
An analyst wants to find systems where users launch interpreters immediately after opening downloaded attachments. What is the most appropriate response for the analyst?
- Risk- and coverage-weighted hunt selection
- Novelty-and-outcome hunt metric
- Execution-chain threat hunt
- Hunt-to-detection engineering handoff
Explanation
Correct Answer
C:Parent-child process lineage and file origin connect initial delivery to potentially malicious execution. The procedure enables testing a document-to-script intrusion pattern.
Incorrect Answers
A:Asset consequence, threat likelihood and visibility gaps can justify investigating rare but damaging behaviors; although useful for prioritizing important hypotheses over common noise, it is not an adequate solution for testing a document-to-script intrusion pattern.
B:New confirmed findings and detection gaps closed are more informative than number of queries executed; the distinction is between measuring the additional value of hunts and testing a document-to-script intrusion pattern; the latter governs this decision.
D:Validated query logic and example events should become tested detection content with owner, limitations and review date; this provides evidence relevant to turning a discovery into sustained monitoring; the remaining need is testing a document-to-script intrusion pattern.
Question 390
A detection content pack is useful at installation but no one checks whether it continues to trigger true cases. What should the team prioritize to meet this objective?
- Consumer adoption and friction analysis
- Longitudinal detection-value tracking
- Comparable validation methodology
- Decision-impact usefulness evaluation
Explanation
Correct Answer
B:Following alert dispositions and missed incidents over time tests continued benefit instead of judging only initial acceptance.
Incorrect Answers
A:Measure recipient understanding, ownership, timing and implementation barriers to explain why technically correct intelligence goes unused; this works toward diagnosing nonadoption of intelligence products; the described problem instead calls for checking continuing value after deployment.
C:Common test data and definitions are required before competing quality scores can be meaningfully ranked; those results can inform building a fair source quality comparison without satisfying the need for checking continuing value after deployment.
D:Useful intelligence should improve a decision or response; publication counts alone do not demonstrate that effect; it is relevant to measuring whether intelligence changes investigative outcomes, but a different capability is needed for checking continuing value after deployment.