ISC · Security and Confidentiality and Privacy · 9 practice questions
Suspected Data Exfiltration: Contain and Preserve Evidence
Contain suspected active data exfiltration immediately while preserving evidence where feasible. Below: follow one compromised server through isolation, evidence capture, and the decision about when to patch.
Try one first
Hint
Think about the first priorities in an incident response: what stops further loss now and what preserves information needed for investigation.
Answer D. Containment and preservation of forensic evidence are the immediate priorities after detecting a suspected compromise. Isolating the affected server limits further data loss while imaging disks and securing logs preserves artifacts investigators need to determine scope, root cause, and attacker persistence. Notification and broader remediation should follow once containment is achieved and the incident scope is better understood.
Why not A: Notification is important and often required, but rushing to notify before containment and verification can impede investigation and may not stop ongoing exfiltration; notifications are typically issued after initial containment and scope assessment or coordinated with the incident response team.
Why not B: Restoring from backup and deploying encryption are valuable recovery and long-term controls, but restoring immediately can destroy evidence needed for investigation and enterprise-wide encryption rollout is not an immediate means to stop ongoing exfiltration.
Why not C: Credential resets can help contain access, which makes this attractive, but performing broad credential changes before imaging systems can destroy forensic artifacts and may not remove all attacker persistence or alternate access vectors; such steps should be coordinated as part of containment after evidence preservation.
Step by step
- Recognize the containment trigger
AlphaCo detects ongoing outbound transfers from its customer database server to an unknown IP after an unfamiliar administrative login. These credible indicators justify immediate containment without waiting to confirm completed exfiltration.
- Isolate without powering off
Isolate the affected server from the network to stop outbound transfers. Keep it powered if safe so the response team can capture volatile evidence.
- Close the compromised access path
Disable the compromised administrative account and terminate its active sessions while preserving available authentication records. A password change alone may leave existing sessions usable.
- Capture volatile evidence promptly
During containment where feasible, capture memory, running processes, active connections, and live logs. Do not leave the attacker connected solely to complete collection.
- Secure persistent evidence
Preserve relevant disk images and system, application, and network logs before destructive changes. Document collection and maintain chain of custody so investigators can reconstruct the server's activity.
- Distinguish containment from disruptive remediation
Once containment is effective, complete volatile capture and preserve relevant disk images and logs before disruptive host patching, restarting, restoring, or rebuilding. If emergency or virtual patching is necessary for immediate containment, do not delay it solely for evidence collection; minimize and document evidence-altering changes.
- Investigate and coordinate next actions
Use the preserved evidence to determine what AlphaCo's attacker accessed and whether data left the server, then guide remediation and recovery. Coordinate notification decisions with legal counsel under applicable requirements; notification does not replace containment.
Key points
- Remember: Cut the attacker's connection, not the server's power, when safe isolation can stop the transfer and preserve memory.
- A disk image does not replace memory capture; shutdown loses volatile evidence.
- Containment and evidence collection can run together; neither requires proof that exfiltration finished.
- Necessary emergency containment can alter evidence; minimize and document those changes.
How the exam traps you
- Wait for completed forensic images before stopping active transfers. Contain immediately and collect evidence concurrently where feasible.
- Power off the server because shutdown stops outbound traffic. Use network isolation while maintaining power when safe, so investigators can capture memory.
- Patch, restart, or restore first because it removes the vulnerability. Preserve evidence before disruptive remediation unless emergency patching is necessary for immediate containment.
Question 2
Hint
Identify what remained exposed after the ACL fix (legacy plaintext backups and long-lived credentials) and choose actions that both stop immediate exposure and prevent recurrence (credential/role changes, encryption enforcement, and preserving evidence).
Answer C. Option C both contains immediate risks (rotate/revoke long-lived credentials that may have been compromised) and addresses systemic causes (use short-lived, least-privilege roles; enforce server-side encryption, preferably SSE-KMS; and enable automated blocking and scanning to catch misconfigurations). Preserving affected backups before re-encrypting or securely isolating/deleting them allows forensic analysis while preventing future recurrence.
Why not A: Tempting because live production data is encrypted, but incorrect here because legacy backups contain plaintext. Monitoring alone doesn't remove the vulnerability of plaintext backups or address compromised long-lived credentials.
Why not B: Tempting because notification is important, but deferring technical containment risks continued exposure. Legal and notification steps should run alongside, not instead of, urgent remediation and forensic triage.
Why not D: Tempting because deletion seems to eliminate risk quickly, but deleting evidence before forensic review can hinder intrusion analysis, regulatory obligations, and root-cause determination. Preserve evidence, then remediate and remediate backups as appropriate.
Question 3
Hint
Prioritize containment that stops ongoing access/exfiltration and preserves evidence; think about isolating the source first rather than broad resets or waiting.
Answer C. Immediate priority is targeted containment: isolate the compromised workstation to stop ongoing outbound connections and preserve volatile evidence by capturing forensic images, and disable only the accounts shown to be compromised to prevent further access. This approach minimizes additional data loss while preserving artifacts needed for investigation. Broader disruptive actions or delays are inappropriate until containment and evidence collection are complete.
Why not A: Although a sweeping reset may seem defensive, it is highly disruptive and may not stop active exfil if the infected host remains connected; it also risks interfering with evidence collection. Targeted account disabling after isolating the host is a better immediate step.
Why not B: Reimaging before capturing forensic images destroys volatile evidence and may not halt ongoing exfil until the host is isolated. Rebuilds are appropriate after containment and evidence collection.
Why not D: Delaying containment when there are successful privileged logins and persistent outbound connections increases the risk of further data loss. Immediate isolation and evidence capture should proceed while the investigation continues.
Question 4
Hint
Focus on whether public exploit code plus active indicators make ongoing exploitation likely and whether the IR plan permits immediate action.
Answer D. A critical vulnerability with publicly available exploit code plus monitoring indicators (unusual outbound connections and automated exploit attempts) makes ongoing exploitation likely. Acme's IR plan explicitly authorizes immediate containment; the team should act to stop likely exfiltration while preserving evidence (forensic snapshots/log capture) rather than waiting for definitive proof.
Why not A: This choice appeals because legal and communication coordination is important. It is incorrect because notifying external parties before controlling an active attack can exacerbate harm; containment and investigation should proceed first, with notifications following once the immediate threat is controlled and facts are known.
Why not B: This is tempting because preserving forensic integrity is important, and staff may worry about altering evidence. It fails because waiting for confirmed exfiltration risks continued compromise and data loss; evidence preservation steps should be taken concurrently with immediate containment, not used to justify delay.
Why not C: Candidates may prefer change-control discipline to avoid operational risk from emergency patches. However, a critical, publicly exploitable vulnerability with active indicators requires urgent mitigation; deferring remediation until the normal window likely permits attackers to succeed in the meantime.
Question 5
Hint
Identify the action that most quickly stops ongoing access and protects logs/evidence; remediation and public communication follow containment.
Answer B. The immediate priority is containment and evidence preservation. Resetting credentials, revoking active sessions, and temporarily blocking external access quickly prevents further unauthorized use of valid credentials while keeping logs intact for forensic analysis. Patching, restoring from backups, and communicating externally are important follow-ups after the incident is contained and the scope is understood.
Why not A: Patching fixes a known vulnerability but may not terminate an active session or immediately stop an attacker who already has valid credentials; it can also change system state and hinder forensics. Containment (credential resets and isolation) is faster for stopping ongoing access.
Why not C: Restoration and adding MFA are valuable remediation steps, but restoring immediately can destroy forensic artifacts and is time-consuming. It also doesn't quickly revoke the attacker's current access when valid credentials are already in use.
Why not D: External notification is important once the scope is known and required by policy or law, but premature public communication before containing the incident and confirming what was affected can cause unnecessary harm. Immediate focus should be containment and investigation.
Question 6
Hint
Focus on what event creates a credible ongoing risk to confidentiality, not on what would prove the breach with certainty.
Answer D. Immediate containment is appropriate because the facts show a credible and ongoing risk: unusual foreign-source authentication, sensitive-table queries, and an attempted bulk export, while the account still has live access. The stem explicitly permits preserving logs and taking a snapshot first, so evidence can be retained before disabling the account. Delaying containment increases the chance of successful exfiltration or data alteration.
Why not A: Valid credentials can be misused if compromised, and unusual access patterns plus a bulk-export attempt outweigh the mere possibility of legitimate activity. Contacting the vendor is reasonable but should not delay containment; investigation and vendor outreach can follow after access is restricted.
Why not B: This distractor appeals to a forensics-first impulse, but it is wrong because proof of completed exfiltration is not required to trigger containment when there is a credible ongoing compromise. Waiting with the account active preserves attacker capability and raises the risk of further disclosure.
Why not C: This option tries to treat the blocked export as possible noise. It is wrong because a single event combined with unusual geography and targeted queries against sensitive tables is a strong indicator of compromise; requiring a repeat attempt improperly raises the response threshold.
Question 7
Hint
Choose the action that immediately eliminates public accessibility while keeping logs and evidence intact for investigation.
Answer D. Primary containment is removing public-read access so no further public downloads occur. At the same time, preserve access logs and capture a forensic snapshot of the bucket to retain evidence for scoping and attribution. Apply least-privilege IAM to reduce recurrence, then proceed to full investigation and notifications once scope is known.
Why not A: Tempting because disclosure notification and remediation are often required by policy and regulators; candidates may think immediate transparency is best. It is wrong as the first step because notifying before containing the exposure can allow continued leakage and may complicate forensic collection; containment should precede broad notifications.
Why not B: Tempting because attackers often use stolen credentials, and credential rotation can block unauthorized access. It is wrong here because the root cause is a public-read misconfiguration, so wide credential invalidation is disruptive, unlikely to stop public reads, and can interfere with forensic continuity if done prematurely.
Why not C: Tempting because encryption is a clear security control. It is wrong as the immediate action because encryption at rest does not retroactively prevent files already made publicly readable from being downloaded and may take time to implement; it does not immediately stop ongoing public access.
Question 8
Hint
Ask whether the thief could decrypt data offline (were keys or cached credentials on the device?) and prioritize containment and evidence collection before broad notification or disruptive enterprise-wide changes.
Answer B. The immediate priorities are containment, evidence preservation, and determining scope: preserve forensic artifacts, attempt a remote wipe when possible, and actively monitor logs for signs of unauthorized access. Because encryption keys were not stored on the device and there is no log evidence of compromise, mass customer notification is not automatically required; notification decisions should follow the documented risk assessment and applicable breach-notification laws.
Why not A: Tempting because stolen PII often leads organizations to notify and remediate quickly to protect individuals. It is wrong here because the data at rest remained encrypted and there is no evidence of key or credential compromise; immediate mass notification and costly remediation before a risk assessment could cause unnecessary alarm and expense.
Why not C: Tempting as an aggressive way to eliminate even theoretical exposure. It is wrong because key rotation and enterprise-wide credential resets are highly disruptive and are typically reserved for confirmed compromise of keys or authentication systems; with no evidence of key compromise, this response is disproportionate and operationally harmful.
Why not D: Tempting because regulatory reporting and law-enforcement involvement are common in breach playbooks. It is wrong as stated because the '72-hour' timeframe is specific to some jurisdictions (e.g., GDPR) and reporting obligations vary; immediate reporting before containment and a preliminary assessment is premature and could hamper forensic work or lead to unnecessary disclosures.
Question 9
Hint
Separate 'containment/eradication of an affected host' (short-term actions) from 'organization‑wide mitigation' (root-cause analysis, patching other systems, resetting shared credentials, and network-wide scanning).
Answer A. The team contained and eradicated the malware on Srv-01 (isolation, removal, restore, and host patching) and logs show no lateral movement, so the immediate incident was contained. However, mitigation requires addressing the root cause and organization-wide risks, those were not addressed (no forensic analysis, no network-wide scans, shared credentials not reset, other servers unpatched). Therefore mitigation is incomplete and further remediation is required.
Why not B: Tempting because the team isolated the host and removed malware, which are key incident-response steps. It is wrong because organizational mitigation also requires root-cause analysis and addressing shared/systemic vulnerabilities; those actions were not performed.
Why not C: Tempting because the team did not scan all systems, leaving uncertainty. It is wrong because the stem explicitly states network logs and endpoint detection showed no evidence of lateral movement or abnormal activity on other hosts, supporting containment of the discovered infection.
Why not D: Tempting for students who focus on the final 'incident closed' phrase and infer benign activity. It is wrong because malware was identified and removed and there were malicious outbound connections, so this was not a benign false positive.
Common questions
What is the first response to suspected data exfiltration?
Contain the active access immediately and preserve evidence concurrently where feasible. Credible indicators of ongoing compromise are enough; do not wait for proof that data left.
Should you shut down a compromised server?
If network isolation safely stops the transfer, keep the server powered for volatile evidence capture. Powering off loses memory and live session information.
Should you patch before collecting forensic evidence?
After effective containment, preserve evidence before disruptive host patching or restarting. If emergency or virtual patching is necessary to contain exploitation, do not delay it solely for evidence collection; minimize and document changes.
Watch it solved
A full CPA ISC task-based simulation on Mitigation, worked step by step.
IT Risk-to-Control Mapping | CPA ISC Simulation (TBS) on YouTube
Practice ISC like the real exam
The free ChatCPA simulator: real exam layout, timed testlets, starting with a question on this topic. No account needed to start.
Open the free simulator →More on Mitigation
- Stolen Credentials: When MFA Is the Best Mitigation
- Full-Disk Encryption on Laptops: Protection and Limits
- Compensating Controls for Unpatched Vulnerabilities
- Vendor Privileged Access: Standing Rights vs. Time-Bound Access
- Least Privilege and Masking for PII: CPA ISC
- Risk Responses: Mitigation, Transfer, Acceptance, Avoidance
Questions from the ChatCPA bank of 17,658 CPA exam questions, each with a written reason for every wrong answer. ChatCPA is built by Nicholas Miller, CPA (Oregon #14907). How these pages are made. Spot an error? Tell us.