Question 60
What verification result would best support closure of a finding about indirect prompt injection in AI web assistant?
- Test documents with embedded instructions cannot make the assistant invoke unauthorized tools or disclose protected data.
- Uploaded polyglot and script-like test files cannot execute or be served with dangerous content types.
- API inspection confirms regular users receive no hidden privileged properties in JSON responses.
- Repeated-redemption test cases cannot exceed the intended one-credit business rule.
Explanation
Correct Answer
A:Indirect prompt injection places adversarial instructions in lower-trust content that a language model may incorrectly treat as authority; the outcome tests this precise property in the production control. It relates to the original observation that a web assistant begins following instructions embedded in a retrieved document instead of the user's authorized request.
Incorrect Answers
B:Uploads can create script, stored-XSS, or malware exposure when the server trusts names and client media types; that evidence is relevant when an avatar service accepts an arbitrary extension and serves the uploaded file with an active browser content type, but does not establish closure of the indirect prompt injection in AI web assistant exposure.
C:Sensitive fields remain exposed when authorization and output filtering occur only in the browser; that evidence is relevant when a customer profile endpoint returns staff-only flags and internal identifiers even though the UI hides them, but does not establish closure of the indirect prompt injection in AI web assistant exposure.
D:Logic flaws exploit allowed operations in an unintended sequence or frequency, beyond ordinary injection filters; that evidence is relevant when a discount checkout allows repeated redemption of a one-time credit without breaking input validation, but does not establish closure of the indirect prompt injection in AI web assistant exposure.
Question 61
A single approved remediation review covers 2 separately named requirements: (1) IoT firmware authenticity; (2) IoT factory credentials. Which audit outcomes would directly demonstrate protection for those areas? Choose TWO.
- Inventory tests show factory logins rejected and approved maintenance identities individually tracked.
- Firewall flow review shows only approved historian exchange paths cross the enterprise and OT boundary.
- Guest and office tests cannot reach the PLC management service without authorized OT access.
- Configuration and network checks prove ordinary plant clients cannot reach the protected admin page.
- Tampered firmware packages fail signature validation while approved updates install normally.
Explanation
Correct Answers
A:This verifies the IoT factory credentials component alongside IoT firmware authenticity in the combined audit: factory credentials create predictable management access if they remain active after device commissioning. For the IoT firmware authenticity audit set, the relevant observed case is that multiple smart sensors share an unchanged default administrative password across an entire facility.
E:This verifies the IoT firmware authenticity component alongside IoT firmware authenticity in the combined audit: unsigned update paths permit modified firmware to be accepted as if it were trusted vendor software. For the IoT firmware authenticity audit set, the relevant observed case is that a gateway installs downloaded firmware without checking that it came from an approved signer.
Incorrect Answers
B:An industrial DMZ mediates data exchange while reducing direct enterprise-to-control-system reachability; this result is evidence about industrial dmz separation when a historian sends process data to enterprise reporting systems through unrestricted two-way connections, not a direct closure test for the named IoT factory credentials, IoT firmware authenticity requirements.
C:PLC control interfaces affect physical processes and must be shielded from unauthorized enterprise traffic; this result is evidence about PLC control network exposure when a programmable logic controller accepts management connections from the office guest network, not a direct closure test for the named IoT factory credentials, IoT firmware authenticity requirements.
D:Embedded management interfaces may retain obsolete defaults and limited authentication controls for years; this result is evidence about embedded web ui hardening when a connected controller exposes a vendor maintenance page with unchanged factory settings over the plant network, not a direct closure test for the named IoT factory credentials, IoT firmware authenticity requirements.
Question 62
The owner says HSTS deployment has been addressed. Which recorded outcome should the independent tester seek?
- Request logs report the expected client address and reject spoofed forwarding-header values.
- Independent handshakes fail for obsolete protocols and succeed for intended modern clients.
- Browser tests upgrade subsequent navigation to HTTPS without exposing the session to a cleartext retry.
- Process identity and permission tests show the worker cannot change privileged host settings.
Explanation
Correct Answer
C:HSTS directs compliant browsers to use HTTPS for the host after receiving a valid secure policy; the outcome tests this precise property in the production control. It relates to the original observation that a site supports HTTPS but returning visitors sometimes attempt insecure HTTP connections first.
Incorrect Answers
A:Proxy metadata is trustworthy only when inserted or overwritten by a controlled trusted reverse proxy; that evidence is relevant when a backend trusts every client-provided forwarding header as the true source IP address, but does not establish closure of the HSTS deployment exposure.
B:Legacy TLS versions and cipher suites may not provide current confidentiality and integrity guarantees; that evidence is relevant when a public HTTPS service negotiates an obsolete transport protocol with weak cryptographic protections, but does not establish closure of the HSTS deployment exposure.
D:Excessive server process rights turn an application compromise into a host-level compromise; that evidence is relevant when a compromised web worker could alter operating-system security settings because it runs as an administrator, but does not establish closure of the HSTS deployment exposure.
Question 63
The assessor flags this situation for follow-up: A legitimate mobile application package appears modified after download from an unofficial distribution source. Select the most appropriate action.
- Store secrets in platform-protected key storage and avoid persistent sensitive debug output.
- Enforce platform certificate trust checks and appropriate endpoint identity validation.
- Distribute signed releases through approved channels and verify publisher provenance.
- Use posture checks and limit sensitive access from noncompliant devices without relying solely on root detection.
Explanation
Correct Answer
C:App signatures help platforms verify package integrity and publisher identity under their trust model; this is the specific exposure or dependency that the selected action addresses. After treatment, the assessor should confirm that package verification rejects altered binaries whose signing identity or integrity is invalid.
Incorrect Answers
A:Locally stored credentials are exposed if applications use unsafe storage locations or logging behavior; the proposed safeguard addresses a case where a mobile app stores persistent authentication tokens as plaintext in a world-readable debug file, rather than the mobile application signing finding described here.
B:Disabling TLS certificate validation enables impersonation even when the connection uses encryption; the proposed safeguard addresses a case where an application accepts any server certificate while connecting to its financial API, rather than the mobile application signing finding described here.
D:Rooted or jailbroken devices may bypass normal sandbox boundaries and weaken enterprise assurance; the proposed safeguard addresses a case where a managed phone has modified its platform trust controls and can access protected application storage, rather than the mobile application signing finding described here.
Question 64
A planned improvement must address the following: A gateway installs downloaded firmware without checking that it came from an approved signer. Which control best targets that requirement?
- Use unique per-device credentials or managed identities and disable unused default accounts.
- Use supported DTLS or equivalent secure CoAP protection for sensitive telemetry.
- Require cryptographically verified updates and a secure boot chain where supported.
- Inventory firmware versions and establish supported staged updates with rollback planning.
Explanation
Correct Answer
C:Unsigned update paths permit modified firmware to be accepted as if it were trusted vendor software; this is the specific exposure or dependency that the selected action addresses. After treatment, the assessor should confirm that tampered firmware packages fail signature validation while approved updates install normally.
Incorrect Answers
A:Factory credentials create predictable management access if they remain active after device commissioning; the proposed safeguard addresses a case where multiple smart sensors share an unchanged default administrative password across an entire facility, rather than the IoT firmware authenticity finding described here.
B:CoAP traffic needs an appropriate protected transport or object security design when data confidentiality matters; the proposed safeguard addresses a case where a constrained sensor sends sensitive readings over unprotected datagrams to an external aggregator, rather than the IoT firmware authenticity finding described here.
D:Unmanaged update processes leave persistent known weaknesses in long-lived embedded deployments; the proposed safeguard addresses a case where deployed devices receive security fixes inconsistently because no operator owns their update policy, rather than the IoT firmware authenticity finding described here.
Question 65
A client requests practical treatment for this condition: A defender examines domain-published text identifying systems allowed to send mail for the domain. Which measure is the closest match?
- Review PTR naming for unintended system details and maintain accurate operational records.
- Permit zone replication only to designated secondary servers using appropriate authorization.
- Publish a scoped sender policy and remove unauthorized sending infrastructure from the allowed set.
- Rotate protected signing keys and publish the matching selector record in DNS.
Explanation
Correct Answer
C:SPF is commonly expressed as a DNS TXT policy and describes authorized outbound mail sources; this is the specific exposure or dependency that the selected action addresses. After treatment, the assessor should confirm that DNS lookup and controlled mail tests show authorized senders align with the published sender policy.
Incorrect Answers
A:PTR records provide reverse address-to-name mapping, which may not mirror forward DNS perfectly; the proposed safeguard addresses a case where a network analyst wants the published hostname associated with an IP address rather than its forward address, rather than the DNS TXT policy records finding described here.
B:Zone transfers are intended for authorized DNS replication, not unrestricted external inventory disclosure; the proposed safeguard addresses a case where an unauthorized client can obtain a broad copy of authoritative DNS zone contents in a single transaction, rather than the DNS TXT policy records finding described here.
D:DKIM binds selected message headers and body content to a signing domain using a published public key; the proposed safeguard addresses a case where an outbound message carries a domain-associated cryptographic signature that receiving mail servers can verify, rather than the DNS TXT policy records finding described here.
Question 66
A remediation ticket for sandbox behavioral analysis needs technical proof before closure. Which test result fits?
- Startup comparisons show only approved persistence entries after remediation and reboot.
- Proxy records show recurring suspicious destinations blocked and the affected endpoint isolated.
- Process trees show suspicious interpreter invocations alerted or blocked despite no new disk executable.
- Sandbox reports capture observed behavior without connections into protected production systems.
Explanation
Correct Answer
D:An isolated analysis environment records process, file, registry, and network behavior with containment; the outcome tests this precise property in the production control. It relates to the original observation that a defender needs to observe a suspicious file's effects without running it on a production workstation.
Incorrect Answers
A:Autorun entries can preserve malicious startup behavior without relying on the original launch process; that evidence is relevant when an unapproved program restarts after each user login because a startup registration remains present, but does not establish closure of the sandbox behavioral analysis exposure.
B:Beaconing can reveal automated communications from compromised endpoints to a remote controller; that evidence is relevant when a workstation periodically contacts an unusual external host with similar request sizes at regular intervals, but does not establish closure of the sandbox behavioral analysis exposure.
C:Fileless malware can abuse built-in interpreters and memory execution to avoid simple file-signature checks; that evidence is relevant when a suspicious task launches code in memory through a trusted interpreter without dropping a conventional binary, but does not establish closure of the sandbox behavioral analysis exposure.
Question 67
After applying safeguards for NTP management query, which observable result most directly establishes that they work?
- Directory audits associate each service principal with an approved owner and low-privilege account.
- Unauthenticated listing requests cannot enumerate private directories on the file-transfer service.
- Public resolver queries show only intentionally exposed records while internal records stay private.
- External clients receive time service as intended but cannot retrieve protected management details.
Explanation
Correct Answer
D:Some NTP management capabilities expose peer and status data unrelated to ordinary time service; the outcome tests this precise property in the production control. It relates to the original observation that a time server responds to management requests with details beyond simple clock synchronization.
Incorrect Answers
A:Service principal names map services to accounts and can indicate privileged service account exposure; that evidence is relevant when a directory assessment reveals service principal names linked to high-value application identities, but does not establish closure of the NTP management query exposure.
B:Anonymous FTP can publish files and names without identity verification or transport encryption; that evidence is relevant when an old file-transfer service exposes a directory listing to a user without authentication, but does not establish closure of the NTP management query exposure.
C:DNS records may expose system naming and topology when public and internal views are mixed; that evidence is relevant when an external reviewer learns internal-looking DNS names from a public-facing name service, but does not establish closure of the NTP management query exposure.
Question 68
A single approved remediation review covers 3 separately named requirements: (1) Known exploited vulnerability catalog; (2) Scanner false negative; (3) AI-assisted vulnerability triage. Which audit outcomes would directly demonstrate protection for those areas? Choose THREE.
- The corrected assessment detects the known affected component and documents the previous coverage gap.
- The remediation ticket cites the relevant catalog entry and verified affected asset software.
- An updated software bill of materials and build scan no longer list the vulnerable dependency version.
- Prioritization records include the current probability estimate and its date, not a fixed severity substitution.
- The risk register shows explicit approver, expiration, interim safeguards, and reassessment evidence.
- The finalized ticket cites validated software evidence, exposure, and business impact rather than the model estimate.
Explanation
Correct Answers
A:This verifies the scanner false negative component alongside known exploited vulnerability catalog in the combined audit: a false negative is a missed real exposure caused by blind spots, permissions, coverage, or detection limits. For the known exploited vulnerability catalog audit set, the relevant observed case is that an independently confirmed vulnerable component is absent from the vulnerability scanner's findings.
B:This verifies the known exploited vulnerability catalog component alongside known exploited vulnerability catalog in the combined audit: a known-exploited catalog identifies vulnerabilities with evidence of exploitation, not just theoretical feasibility. For the known exploited vulnerability catalog audit set, the relevant observed case is that an operations team needs a reliable signal that a CVE has been observed in active exploitation.
F:This verifies the AI-assisted vulnerability triage component alongside known exploited vulnerability catalog in the combined audit: probabilistic AI summaries can assist prioritization but do not replace authenticated evidence or analyst validation. For the known exploited vulnerability catalog audit set, the relevant observed case is that an automated AI triage report labels a service vulnerable despite missing package evidence and system-owner context.
Incorrect Answers
C:Third-party transitive dependencies can introduce vulnerabilities independently of directly written code; this result is evidence about software dependency risk when a web application bundles a library with a known flaw even though the application code was not changed, not a direct closure test for the named scanner false negative, known exploited vulnerability catalog, AI-assisted vulnerability triage requirements.
D:EPSS estimates the probability of exploitation in the wild over a defined period rather than business impact; this result is evidence about EPSS probability estimation when two vulnerabilities have similar severity, but a forecast model assigns materially different exploitation likelihoods, not a direct closure test for the named scanner false negative, known exploited vulnerability catalog, AI-assisted vulnerability triage requirements.
E:Risk acceptance needs accountable approval, defined scope, expiry, and reassessment triggers; this result is evidence about time-limited risk acceptance when a business owner accepts an unresolved vulnerability with no review date or documented compensating control, not a direct closure test for the named scanner false negative, known exploited vulnerability catalog, AI-assisted vulnerability triage requirements.
Question 69
A client-approved assessment includes this finding: A mobile accessory repeatedly requests pairing while advertising an unexpected device identity. What is the most precise security term for it?
- Bluetooth pairing trust
- Evil twin access point
- WPA2-Enterprise authentication
- WPS PIN exposure
Explanation
Correct Answer
A:Bluetooth pairing establishes device trust that should be validated to avoid unintended connections; that mechanism makes this classification more precise than the neighboring terms. The protective follow-up is to use secure pairing modes and verify device identities before authorizing trust relationships, verified when the accessory rejects unknown pairing attempts and retained bonds match approved devices.
Incorrect Answers
B:An evil-twin AP impersonates a trusted network name to attract client associations or credentials; this different classification would fit when a laptop connects to a nearby access point using the company SSID but an unfamiliar certificate and controller, not the evidence pointing to Bluetooth pairing trust in the question.
C:Enterprise WLAN access uses 802.1X and an authentication backend to manage per-identity access; this different classification would fit when a corporate wireless network needs distinct user or device identities instead of one shared password, not the evidence pointing to Bluetooth pairing trust in the question.
D:WPS PIN enrollment can undermine passphrase security on implementations with weak PIN protections; this different classification would fit when a wireless router allows legacy PIN-based onboarding despite a strong Wi-Fi passphrase, not the evidence pointing to Bluetooth pairing trust in the question.
Question 70
Review the following evidence from an approved test: A poster's QR image opens a credential collection page while appearing to link to a visitor check-in. Which concept explains it most directly?
- QR-code phishing
- MFA approval deception
- Voice phishing
- SMS phishing
Explanation
Correct Answer
A:QR-based phishing hides the destination inside an encoded image, making URL inspection less obvious; that mechanism makes this classification more precise than the neighboring terms. The protective follow-up is to use a trusted check-in domain and train users to inspect destinations before entering credentials, verified when test scans confirm users recognize and reject unexpected external login domains.
Incorrect Answers
B:Social pressure can convert an MFA approval into user-assisted account takeover despite a separate factor; this different classification would fit when an attacker instructs a user to approve an unsolicited authenticator prompt as part of a supposed help-desk repair, not the evidence pointing to qr-code phishing in the question.
C:Vishing exploits voice conversations to elicit secrets or persuade users to take unsafe actions; this different classification would fit when a caller posing as technical support asks an employee to read a one-time login code aloud, not the evidence pointing to qr-code phishing in the question.
D:Smishing delivers fraudulent instructions through text messages rather than email or phone calls; this different classification would fit when a text message uses a package-delivery pretext to send a user to an unfamiliar sign-in site, not the evidence pointing to qr-code phishing in the question.