Home SERVICES
All Services Web App Security Network Testing Cloud Security Active Directory Red Team AI Red Teaming
COMPANY
About Us Founder, Arturs Stay Certifications Why Organizations Trust Cyber Security Pentesting FAQ
Process Partners Industries Blog Book a Call
Back to Blog
Vulnerability Response

Zero-Day Vulnerability Triage: An Enterprise Guide for Canada and the US

Enterprise zero-day vulnerability triage for Canada and the USA: exposure, prioritization with CISA KEV, EPSS and SSVC, temporary mitigation, authorized validation and verified remediation, aligned to NIST SP 800-61 Rev. 3 and NIST CSF 2.0.

Zero-Day Vulnerability Response Exposure Triage CISA KEV
↓  Get the Zero-Day Triage Playbook

TL;DR: Executive Brief

  • Confirm exposure first. Build a verified inventory of affected products and versions, internet-facing and management interfaces first. An asset you cannot check is unknown, not safe.
  • Prioritize on evidence. CISA KEV shows known exploitation, EPSS estimates likelihood and SSVC turns them into a decision; add your own exposure and business criticality. CVSS describes severity, not your risk.
  • Mitigate and preserve evidence. Capture logs before they rotate, apply the vendor mitigation or remove the vulnerable path, and verify that the measure actually works.
  • Validate through authorized testing. Narrow, written-authorized testing confirms exploitability, mitigation effectiveness and attack paths, never with untrusted exploit code against production.
  • Escalate suspected compromise. Any indicator of exploitation moves that asset to incident-response command. A patch is not proof that an attacker who got in earlier has left.

Run these steps during a live event with the free Zero-Day Triage Playbook.

A zero-day vulnerability is a security flaw that attackers can exploit before the vendor has released a fix, and it gives every enterprise the same problem on the same morning: a vendor or government agency confirms that attackers are exploiting a flaw in technology you may run, and there is no patch yet, or there is a patch nobody has tested. For organizations in Canada and the United States, the real damage rarely comes from the announcement itself. It comes from the days that follow, when nobody can say with confidence which systems are affected, which are reachable from the internet, and whether any were already exploited.

That uncertainty is what this guide is about. It treats zero-day response as exposure triage: determine exposure quickly, prioritize with evidence, mitigate before the patch where you must, verify that it works, and hand anything that looks like compromise to incident response at once. It is written for CISOs and the vulnerability-management, SOC, incident-response, IT, cloud and application teams who do the work, and it complements our ransomware response and recovery playbook, which starts where this guide stops: once an attacker is inside.

Executive takeaway: The organizations that come through a zero-day well are not the ones with the fastest patch. They are the ones that can answer three questions within hours, with evidence: where do we run it, which instances can an attacker reach, and is there any sign it was used against us?
Zero-Day Triage at a Glance
INTAKE → SCOPE → PRIORITIZE → CONTAIN → VALIDATE → REMEDIATE → VERIFY → LEARN

The operational sequence this guide follows. Scoping, prioritization and containment run in parallel workstreams; any sign of exploitation moves the affected asset to incident response.

Executive Summary

Zero-day response is decision-making under uncertainty; the organizations that handle it well decide in advance how to reduce that uncertainty and who may act on it.

Question Summary
The problemA zero-day forces decisions on shutting, patching, restricting and disclosing before anyone knows whether you are affected, exposed or already compromised.
Why it mattersExploitation now routinely starts before or within days of disclosure, and edge devices such as VPNs, firewalls and gateways are among the most targeted products.
Why leadership should careDowntime, disclosure and incident costs are decided in the first days. Disciplined triage makes those calls defensible instead of guesses.
The recommended approachRun zero-day response as exposure triage: shared status codes, a verified inventory, prioritization by exploitation, exposure and criticality, verified mitigations, an immediate hand-off to incident response at the first indicator, and authorized testing that proves which exposures are real.

Why Zero-Day Vulnerabilities Become Major Business Risks

A zero-day becomes a business risk when it lands on a system that sits on a trust boundary, faces the internet, and cannot be patched or watched as easily as a laptop. That describes many of the products attackers now favor.

Google Threat Intelligence Group's 2025 zero-day review tracked 90 zero-days exploited in the wild in 2025, and 43 of them, 48 percent, affected enterprise technologies, the highest number and share it has recorded. Security and networking products, the VPNs, firewalls and gateways that guard the perimeter, were a large part of that group. Mandiant's M-Trends 2026 estimated the mean time to exploit at minus seven days, meaning that, on average, exploitation began before a patch was available, and found exploits were the most common initial infection vector for the sixth year running, in 32 percent of intrusions. Verizon's 2026 Data Breach Investigations Report reported that exploitation of software flaws, at 31 percent, has overtaken stolen credentials as the leading entry point into breached organizations.

The Canadian Centre for Cyber Security's National Cyber Threat Assessment 2025-2026 adds the Canadian view: threat actors very likely target edge devices because defenders may have limited ability to monitor them, and attacks start within days of disclosure.

IBM's 2026 Cost of a Data Breach Report, conducted by the Ponemon Institute, put the global average cost of a data breach at USD 4.99 million and the Canadian average at a record CA$7.11 million. These are averages across all breach types, not the cost of a zero-day, but they show how one exploited perimeter system can become a multimillion-dollar incident once access turns into data theft, extortion or disruption.

For a CISO, the risk is a decision made blind. For a CIO, it is an emergency change against a business-critical system with no tested rollback. For a board, it is the question that follows every public zero-day: were we affected, and how do we know?

What a Zero-Day Is, and What Active Exploitation Actually Means

A zero-day vulnerability is a flaw that attackers know about, or are exploiting, before the vendor has released a fix. The term describes the defender's position, not the severity. Precise language matters, because most bad decisions during a zero-day come from treating different states as one.

Term What it means What it does not mean
Zero-day vulnerabilityA flaw attackers can use before a fix existsThat you are affected or being attacked
Zero-day exploitWorking code or a technique that uses the flawThat it has been used against you
Zero-day attackUse of a zero-day exploit against a real targetThat your organization was the target
Active exploitationCredible evidence, usually from the vendor or a government agency, of real attacksThat every exposed system has been hit
Known exploited vulnerabilityA CVE listed in the CISA Known Exploited Vulnerabilities (KEV) catalog, which requires a CVE ID, reliable evidence of exploitation in the wild and clear remediation guidanceThat unlisted flaws are safe
Vulnerability exposureA vulnerable version whose vulnerable interface an attacker can reachThat anyone has exploited it
Confirmed compromiseEvidence that an attacker exploited the system or acted through itThat the attacker has been removed by patching
Emergency incident responseInvestigating, containing and recovering from a suspected or confirmed compromiseA penetration test
Authorized penetration testingAuthorized testing of exposure, exploitability and attack pathsForensics, or proof of no compromise
A missing patch is not a compromise, and a patch is not proof of safety. An unpatched system is exposed until it is fixed or protected. A patched system may still have been exploited before the patch, and an attacker who got in earlier may still be there. Treat these as separate questions with separate evidence.

Free download: the Zero-Day Vulnerability Triage Playbook

The article explains the approach; the Playbook is the operational workbook you fill in during a live zero-day event: status codes, checklists, worksheets and report templates, so your team does not build them under pressure. Confirm your email and we send a secure download link.

Exposure status codes First-hour checklist Escalation RACI Exposure inventory KEV / EPSS / SSVC worksheet Evidence preservation Mitigation matrix Patch sequencing Validation worksheet Executive status report 24/48/72-hour tracker

Privacy notice: Double opt-in - we email a confirmation link and the playbook downloads only after you confirm. We use your email to deliver the playbook. Your name, company, job title, phone and LinkedIn URL are stored with your request record and used to confirm that requests come from businesses and to let us know who is requesting our resources; they are not needed to send the file. Requesting the playbook does not subscribe you to marketing: we send marketing communications only if you tick the optional box above, and you can unsubscribe at any time. We never sell your information. To unsubscribe or request deletion, email info@cybersecpentesting.com. See our Privacy Policy.

Defensive and authorized-use resource. The Zero-Day Vulnerability Triage Playbook is provided for defensive security and vulnerability-response purposes on systems you own or operate. Any active security testing it describes requires written authorization and agreed rules of engagement, and must never involve running untrusted exploit code against production systems. It is not legal advice, and it does not replace your incident-response plan or provider.

Why Vulnerability Scanning Alone Is Not Enough

Scanners answer a narrow question, whether a known signature or version string matches, and a zero-day arrives before most scanners can ask it. A scanner is an essential input to triage, not the triage itself.

In the first hours there may be no detection plugin. When one arrives it often relies on version banners that appliances hide, and many edge products cannot be scanned with credentials, so a clean result can mean "not vulnerable" or "not detected". Scanners also cannot say whether this instance, as configured, is exploitable, or what an attacker could reach from it.

Question Vulnerability scanner Inventory and attack-surface discovery SOC telemetry and hunting Authorized validation testing
Do we run the affected product?Partly, if a plugin existsYes, if the inventory is completeSometimes, from trafficConfirms on request
Is the vulnerable interface reachable?From the scanner's placement onlyPartlyPartly, from flow dataYes, from agreed vantage points
Is this instance exploitable as configured?RarelyNoNoYes, by controlled proof
Does the mitigation block the vulnerable path?RarelyNoPartly, if attempts are seenYes, for the known vulnerable path
Was it exploited?NoNoYes, within telemetry coverageNo, this is incident response
What could an attacker reach from it?NoNoNoYes, by attack-path testing

The instances that cause the most trouble are usually the ones nobody listed. Disaster-recovery peers, lab and pre-production appliances, systems inherited through an acquisition, and management interfaces left reachable "temporarily" are where exposure hides. Comparing an independent external view with the inventory, as described in our guide to internal versus external penetration testing, is often the fastest way to find them.

The First Decisions a CISO Must Make

The first decisions are about command and vocabulary, not technology: who leads, what status words everyone uses, what triggers downtime, and when incident response takes over. Settle them in the first thirty minutes.

  1. Name an event commander. One person owns priorities; a scribe records decisions in UTC.
  2. Adopt a shared status vocabulary. Every asset gets one of the six status codes below; nobody reports "we're fine" without one.
  3. Set the downtime threshold. Decide who can take a Tier 0 or Tier 1 service offline to mitigate, and how fast.
  4. Define the incident trigger. Any indicator of exploitation moves that asset to incident-response command.
  5. Control the narrative. No statement about being unaffected until the evidence supports it and counsel has reviewed it.
Code Status Meaning Default handling
S0UnaffectedProduct or version verified absent by direct checkRecord the evidence
S1Affected, not externally exposedVulnerable version present; vulnerable interface not reachable from untrusted networksExpedited patch or mitigation
S2Exposed, exploitation not establishedReachable by an attacker; no indicators found so far, with hunt status and telemetry coverage recordedMitigate now; hunt retrospectively
S3Suspected exploitationIndicators or anomalies consistent with exploitationEscalate to incident response; preserve evidence
S4Confirmed compromiseIncident response confirmed attacker activityIncident response leads
SUUnknownPresence, version or reachability cannot be determinedTreat as at least S2; assign an owner and deadline
Unknown is not safe. An asset you cannot locate or version is SU. An exposed asset whose log retention is shorter than the exposure window stays S2 with the telemetry gap recorded: "no evidence of exploitation" then means "no evidence available", and it is reported that way.

Identifying Affected Products, Exposed Assets and Vulnerable Dependencies

Scoping means turning the vendor's list of affected versions into a list of your instances, each with a verified version, a verified network position and an owner. Verified is the important word: inventories are a starting point, not evidence.

Start with internet-facing and management-plane instances, because that is where exploitation starts. Confirm versions from the management console or the vendor's own check, not only the CMDB. Then widen the search: high-availability and disaster-recovery peers, non-production and lab systems, cloud images and templates that would redeploy the vulnerable build, and partner-hosted or MSP-managed instances.

Zero-days in libraries and embedded components are harder, because the vulnerable code hides inside other products. Software bills of materials for the applications you build, and written statements from suppliers about the products you buy, are the only scalable way through. For third-party and SaaS services, record who you asked, when, and what they answered; a supplier that cannot confirm its status is an open risk, not a closed one. Our analysis of the enterprise supply chain attack surface covers why these dependencies are so often missed.

Where the vulnerable product is an internet-facing appliance or service, an external view of reachability is worth having early, from your own attack-surface tooling or an existing provider. A scheduled external assessment, the core of network penetration testing for the perimeter, then confirms from an attacker's vantage point what is actually exposed rather than what the firewall diagram says.

Prioritizing With KEV, EPSS, Exposure, Exploitability and Business Criticality

Vulnerability prioritization for a zero-day works when each signal answers the question it was designed for. The CISA Known Exploited Vulnerabilities (KEV) catalog tells you exploitation is known. The Exploit Prediction Scoring System (EPSS), maintained by FIRST, estimates the probability that a published CVE will be exploited in the wild in the next 30 days. CVSS describes technical severity. CISA's Stakeholder-Specific Vulnerability Categorization (SSVC) guide combines exploitation status, technical impact, whether exploitation is automatable, and mission and well-being impact into four decisions: Track, Track*, Attend and Act. None of them knows your exposure or your business; you add those.

Signal Question it answers Common mistake
CISA KEVIs exploitation in the wild known and reliable?Treating unlisted flaws as safe; KEV listing can lag exploitation
EPSSHow likely is exploitation activity in the next 30 days?Reading it as impact; FIRST calls it the threat-likelihood component only
CVSSHow severe is the flaw technically?Using the base score as a risk score for your environment
SSVCGiven exploitation, impact, automation and mission, what should we do?Skipping the mission inputs that make it yours
ExposureCan an attacker reach the vulnerable interface?Inferring reachability from diagrams instead of testing it
Business criticalityWhat does this system protect and what does downtime cost?Treating edge security appliances as low-value infrastructure

US federal practice now points the same way. In June 2026, CISA replaced BOD 22-01, the directive that created the KEV catalog, with Binding Operational Directive 26-04, which ties federal remediation urgency to asset exposure, KEV status, whether exploitation is automatable, and technical impact; agencies must begin remediating on its timelines by December 7, 2026, and FedRAMP has aligned its cloud-provider rules to it. The directive binds US Federal Civilian Executive Branch agencies, not private enterprises in Canada or the US. For everyone else, KEV, EPSS and SSVC are recommended industry practice rather than legal requirements, and your obligations come from your sector regulators, privacy law and contracts. The principle still transfers: exposure and exploitation decide urgency.

A per-asset matrix, provided in the playbook as a worksheet, looks like this; calibrate the targets in policy before an event.

Priority Typical conditions Example target for first risk reduction
P1Active exploitation confirmed and the asset is exposed (S2) or unknown (SU) on an internet-facing or management interface; or any suspected exploitation (S3)Hours: mitigate, restrict or isolate in the first operating window
P2Active exploitation and the asset is internal-only (S1) on a Tier 0 or Tier 1 system; or public proof of concept and the asset is exposedExpedited: days, ahead of the normal patch cycle
P3Proof of concept or no known exploitation; internal exposure; Tier 1 or Tier 2Next accelerated patch window
P4No known exploitation; not exposed; Tier 3Normal patch cycle

If an asset matches no row, or more than one, assign the higher of the candidate priorities and record the reason in the decision log. Confirmed compromise (S4) sits outside this matrix: incident response leads.

No score is not a low score. A brand-new zero-day may have no CVE, no EPSS score and no KEV entry for hours or days. When the vendor or a government agency confirms active exploitation, treat exploitation as active regardless of what the scoring feeds say.

The First Hour, and When Incident Response Must Take Command

The first hour is about preserving evidence and building an exposure picture in parallel, so that by the end of it you can mitigate exposed systems and know which assets belong with incident response. The condensed sequence below comes from the playbook, where each item is assigned to a workstream.

First 15 minutes
Open & preserve
  • Open the event; name a commander and a scribe who records decisions in UTC.
  • Capture the vendor advisory, CVE record, KEV entry and Canadian Centre for Cyber Security alert with retrieval times.
  • Export or extend logs for the affected product before rotation removes the exposure window.
  • Do not run public exploit code, reboot appliances before evidence capture, or say you are unaffected.
First 30 minutes
Scope & hunt
  • Build the exposure list, internet-facing and management interfaces first.
  • Verify versions directly where inventory may be stale.
  • Start a retrospective hunt from the earliest plausible exploitation date, not the disclosure date.
  • Engage your incident-response provider if anything looks wrong or you cannot hunt yourself.
First 60 minutes
Decide & contain
  • Decide per exposed asset: mitigate, restrict, isolate or accept with a named approver.
  • Capture evidence before any mitigation that restarts the service, then apply vendor mitigations, most critical first.
  • Enable detections for exploitation attempts and post-exploitation behavior.
  • Issue the first executive status report and set the next update time.

The hand-off rule is simple and should be absolute: as soon as an asset shows indicators of exploitation, it moves to incident-response command. The goal then changes from closing exposure to removing an intrusion, and routine triage actions such as patching, rebooting or testing can destroy the evidence responders need. Patching does not evict an attacker who has already harvested credentials or established persistence. If the event escalates into a full intrusion, the containment, identity and recovery sequencing in our ransomware incident-response guide becomes the next playbook.

This model is consistent with NIST SP 800-61 Rev. 3, published in April 2025, which treats incident response as part of cybersecurity risk management under the NIST Cybersecurity Framework 2.0, and with CISA's incident and vulnerability response playbooks, which describe vulnerability response as a process of identification, evaluation, remediation, and reporting that hands confirmed compromise to incident response.

Coordinating IT, SOC, Incident Response, Legal, Cloud and Business Owners

Zero-day response fails at the seams between teams, so the coordination model has to be explicit: one commander, defined workstreams, named decision owners, and a fixed reporting cadence that gives status counts by code, what changed, and the decisions needed and by when.

Workstream Owns Typical decision owner
CommandPriorities, status, reporting cadence, escalationEvent commander (CISO or delegate)
Asset and exposureInventory, versions, reachability, status codesVulnerability management with asset owners
Detection and huntingRetrospective hunt, new detections, indicator matchingSOC lead
Mitigation and changeEmergency changes, patch sequencing, rollbackIT operations, cloud and platform owners
Incident responseAny asset with indicators; evidence; scoping of intrusionIncident-response lead or provider
Legal, privacy and communicationsNotification analysis, privilege, external statementsCounsel and communications lead
BusinessDowntime and residual-risk decisionsBusiness owner and executive sponsor

Cloud and SaaS deserve their own line. In public cloud the vulnerable component may be yours, the provider's or a marketplace image; for SaaS you depend on the supplier's statement. A provider's fix does not rotate credentials or tokens that may have been exposed in your tenant. Where a zero-day touches cloud identity or workloads, the same validation discipline applies, and our cloud penetration testing work focuses on exactly those identity and configuration paths.

Temporary Mitigations When No Patch Exists

When there is no patch, or the patch cannot yet be applied safely, the goal is to remove the attacker's path to the vulnerable code: apply the vendor's mitigation exactly, disable the vulnerable feature, or take the interface off untrusted networks. Then verify that the measure works.

Option When it fits Watch out for
Vendor mitigation or workaroundThe vendor has published onePartial application; later updates that silently revert it
Disable the vulnerable featureThe feature is not business-criticalDependencies nobody documented
Restrict the interface (allow-list, ACL, VPN or zero-trust access)The interface does not need public reachabilityForgotten paths: DR peers, IPv6, secondary addresses
WAF or IPS virtual patchAs a supplement while a stronger control is put in placeBypasses are common; never the only control for P1 assets
Isolate or fail overNo other control is effective, or exploitation is suspectedBusiness impact; prefer isolation over power-off
Revoke sessions, tokens or credentialsThe flaw exposes sessions or secrets, or the vendor requires it after patchingFor S3 and S4 assets, coordinate timing with the incident-response lead; otherwise follow vendor instructions without delay

Recent history shows why verification matters. With the Citrix Bleed vulnerability, CVE-2023-4966, NetScaler's guidance told customers to remove active and persistent sessions after upgrading, because hijacked sessions survived the patch. With the Ivanti Connect Secure zero-days of January 2024, a vendor mitigation was available before patches, and CISA later reported in joint advisory AA24-060B that the vendor's integrity checker could fail to detect compromise. And in the MOVEit Transfer campaign described in CISA and FBI advisory AA23-158A, a previously unknown SQL injection flaw was used for large-scale data theft before most organizations could act. Each case teaches the same rule: mitigation and patching reduce future exploitation; they do not answer whether exploitation already happened.

NIST SP 800-40 Rev. 4 frames patching as preventive maintenance and asks organizations to choose and record a risk response (accept, mitigate, transfer or avoid) for each vulnerability; when no fix exists, mitigation has to come from other controls. A temporary mitigation without an owner and a review date tends to become permanent.

How Targeted Penetration Testing Validates Exploitability and Exposure

Targeted penetration testing answers the questions scanners and inventories cannot: is this instance actually exploitable in your configuration, does the mitigation really block the vulnerable path, and is the interface reachable from where an attacker would stand.

Done properly, it is narrow, authorized and safe: written authorization from the asset owner and an accountable executive, rules of engagement listing permitted and prohibited techniques, SOC deconfliction, and confirmation from incident response that no target is under investigation. It never runs untrusted public exploit code against production; proof of exploitability, where needed, happens in a production-equivalent replica or through agreed non-destructive checks.

Area What it can establish What it cannot establish
ExposureWhether the vulnerable interface is reachable from defined vantage pointsEvery path an attacker might find in the future
ExploitabilityWhether a specific instance, as configured, is exploitable, by controlled proofThat an untested sibling instance behaves the same
MitigationWhether the mitigation blocks the known vulnerable pathThat a novel bypass will never be found
CompromiseNothing directly; anomalies are passed to incident responseWhether exploitation occurred: incident response and forensics work
Business impactWhat an attacker could reach from the asset, through attack-path testingThe financial cost of a hypothetical breach

For decision-makers, a validated exploitable instance justifies emergency downtime, while a validated mitigation can justify waiting for a scheduled patch window on a revenue-critical system. When the vulnerable component sits inside a customer-facing web application or API, web application penetration testing can confirm whether the vulnerable code path is reachable through your application's actual routes and controls.

When the vulnerable product is unusual, an OT gateway, a niche file-transfer platform, an acquired company's stack, a custom penetration test scoped to that one exposure is usually more useful than a broad assessment.

How Attack-Path Testing Identifies Downstream Business Risk

Attack-path testing asks what an attacker could do next from the vulnerable system, which is where most of the business impact of a zero-day actually sits. An exploited VPN appliance is a problem; an exploited VPN appliance that stores a directory bind account with broad rights is a domain-wide problem.

Edge appliances and file-transfer systems commonly hold credentials, sessions, keys and network position, and attack-path testing follows them into identity, privileged groups, backup and management planes. Our enterprise guide to Active Directory attack paths explains how those identity paths are mapped and proven, and network segmentation testing shows how to confirm whether the segment an appliance sits in actually contains it.

Starting point What the attacker may gain Downstream risk to test
VPN or remote-access applianceActive sessions, cached credentials, directory bind accountDomain privilege escalation; access to internal applications
Managed file-transfer serverStored files, service accounts, database credentialsRegulated-data theft; lateral movement to databases
Firewall or gateway management planeConfiguration, VPN keys, routing controlTraffic interception; pivot into protected segments
Internet-facing application serverApplication secrets, cloud access keysCloud control-plane access; data store exposure

When identity is in the path, testing whether the directory and privileged-access controls hold is often the most valuable follow-up to a zero-day. That is the focus of Active Directory security testing: showing whether credentials taken from a perimeter device can become control of the domain.

Post-Patch Retesting and Compensating-Control Validation

A change is not risk reduction until it is verified against the original vulnerable path. Patches miss a cluster node, mitigations are reverted by later updates, and compensating controls block the published payload but not a trivial variant.

Verification has a short, repeatable checklist: confirm the fixed version from the management plane and from the service itself; re-check the vulnerable path; test access restrictions from an external vantage point; confirm detections fire with a controlled test; rotate sessions or secrets where the vendor requires it; confirm images and templates no longer carry the vulnerable build; and have incident response review the retrospective hunt for the exposure window. Close the event only when every asset has a final status and every remaining risk has an accountable owner.

Retesting converts "we applied the patch" into "we verified the exposure is closed", which is what auditors and boards need. Where retesting is part of an engagement, its scope, timing and terms are set in the approved proposal.

Regulatory, Insurance and Operational Considerations for Canadian and US Enterprises

Exposure to a zero-day is rarely reportable on its own, but suspected or confirmed compromise can start legal clocks quickly, so triage records need to be good enough for counsel to decide on. This section is general information, not legal advice.

Jurisdiction or regime Who it applies to What it requires (summary)
Canada: PIPEDAPrivate-sector organizations subject to PIPEDAReport breaches of security safeguards that create a real risk of significant harm to the Privacy Commissioner, notify affected individuals, and keep records of all breaches for two years (PIPEDA breach-reporting requirements)
Canada: OSFIFederally regulated financial institutionsReport qualifying technology or cyber incidents to OSFI within 24 hours, or sooner if possible (incident-reporting advisory); Guideline B-13 sets technology and cyber risk expectations
Canada: provincial lawsOrganizations subject to Quebec's Law 25, Alberta PIPA or provincial health-privacy actsProvincial incident reporting, notification and register obligations that can apply alongside or instead of PIPEDA
US: SECPublic companies registered with the SECDisclose a material cybersecurity incident on Form 8-K, generally within four business days of determining materiality (SEC cybersecurity disclosure rules)
US: New York DFSCovered financial entitiesNotify the superintendent no later than 72 hours after determining a cybersecurity incident occurred (23 NYCRR Part 500)
US: sector and state rulesHealth, critical infrastructure, state residents' dataHIPAA, state breach laws and sector regulators add triggers, including the federal banking agencies' 36-hour notification rule for banking organizations; the federal CIRCIA reporting rule was not yet in effect in November 2026

Two operational points matter as much as the rules. Cyber-insurance policies often set their own notification terms and may require approved responders or counsel, so check the policy before you need it. And your triage record decides how quickly counsel can answer the question every regulator asks: when did you know, and what did you know? A timestamped decision log and status codes make that answer defensible.

Testing provides evidence for these processes; it does not make an organization compliant, and auditors and regulators decide compliance.

An Illustrative Enterprise Zero-Day Scenario

The following is an illustrative composite built from common patterns, not an account of any client, incident or outcome.

A mid-sized insurer with offices in Toronto and Chicago learns that its remote-access appliance has a vendor-confirmed, actively exploited zero-day. A mitigation is available; a patch is promised within days.

  • Hour 0 to 1. An event commander is named and logs are exported. The inventory lists two internet-facing appliances, marked S2 (hunt pending) and mitigated under a pre-approved emergency change.
  • Hour 2 to 6. External attack-surface data reveals a third instance: a disaster-recovery appliance missing from the CMDB. It moves from SU to S2 and is mitigated. The hunt finds no indicators on the production pair, but the DR unit keeps only seven days of logs, so it remains S2 with the telemetry gap recorded (retention shorter than the exposure window) and the incident-response provider is consulted.
  • Day 1 to 2. Authorized external validation shows the mitigation holds on two appliances but not on the DR unit, where a configuration sync had partly reverted it; it is re-applied and re-verified. Attack-path review finds the appliances use an over-privileged directory bind account.
  • Day 3. The patch is applied in sequence, versions verified, sessions reset and the bind account reduced to least privilege. Counsel reviews the record, and the DR telemetry gap is recorded as an owned residual risk.

The sequence relied on a status vocabulary, an independent view of exposure, verified mitigation, and a clean line between exposure and compromise. The playbook makes each step repeatable.

What to Expect From a Zero-Day Exposure Validation Engagement

A zero-day exposure validation engagement is a narrowly scoped, authorized test of your exposure to a specific vulnerability: whether the affected instances are reachable, whether they are exploitable as configured, whether mitigations hold, and what an attacker could reach from them. It is not incident response, forensics, or a substitute for your vendor's patch.

Who performs the workA named principal consultant (CREST CRPT, OSCP, OSEP, CRTO), not an outsourced or rotating team. The person who scopes the work performs it and writes the report.
Access requiredWritten authorization for named assets and a time window, your exposure list, and a production-equivalent replica or agreed non-destructive method where proof of exploitability is needed.
Production impactDesigned to be non-destructive. In production, only read-only and non-intrusive checks are run; anything with possible service impact happens in a replica or is explicitly agreed in the rules of engagement. No untrusted exploit code is run against production.
TimelineSet by scope and confirmed in the approved proposal. Booking a scoping call does not start an emergency or incident-response engagement.
How findings are validatedEach result (exploitable, not exploitable, mitigated, inconclusive) is reproduced and evidenced, with method and vantage point recorded.
DeliverablesA written report with an executive summary and technical findings, the validation log and attack-path results, followed by a debrief where agreed.
RetestingRetesting and remediation validation are defined in each approved engagement proposal.
Not includedIncident response, digital forensics, 24/7 monitoring, or legal and regulatory advice. If you suspect compromise now, engage your incident-response provider first.

What You'll Receive in the Zero-Day Triage Playbook

The playbook complements this guide rather than repeating it. Alongside exposure status codes, a printable escalation card and a sourced references section, it contains fifteen worksheets:

  • A. Event intake record and source-reliability log
  • B. First 15, 30 and 60-minute checklist by workstream
  • C. Escalation thresholds and RACI matrix
  • D. Affected product, version, dependency and supplier tracking
  • E. External and internal exposure inventory
  • F. CISA KEV, EPSS and SSVC prioritization worksheet
  • G. Business-criticality classification
  • H. Exploitation indicators, evidence preservation and chain of custody
  • I. Containment and temporary-mitigation decision matrix
  • J. Emergency change control and patch sequencing
  • K. Authorized exploitability and attack-path validation worksheet
  • L. Remediation and compensating-control verification
  • M. Executive and CISO status report template
  • N. 24, 48 and 72-hour tracker and decision log
  • O. Lessons learned and exposure-reduction checklist
Get the playbook: Use the download form above to receive your secure copy. Want evidence on whether a specific exposure is real? Book a Zero-Day Exposure Scoping Call to discuss authorized validation for enterprises in Canada and the United States.

Frequently Asked Questions

What is a zero-day vulnerability?

A zero-day vulnerability is a security flaw that attackers know about, or are exploiting, before the vendor has released a fix, so defenders have had zero days to patch. The term describes the defender's position, not the severity. Once a patch exists the flaw is often called an n-day, and it stays dangerous until systems are patched.

Does a zero-day in a product we use mean we have been compromised?

No. A vulnerable version is exposure, not compromise. Compromise requires evidence that an attacker exploited the flaw or acted through the system. If logs are missing, or retention is shorter than the exposure window, you cannot say either way, and that gap should be treated as risk until it is resolved.

How should we prioritize a zero-day with CISA KEV, EPSS and CVSS?

Use each for its own question. KEV confirms known exploitation, EPSS estimates the likelihood of exploitation activity in the next 30 days, CVSS describes technical severity, and SSVC turns these into a decision. Then apply your exposure and business criticality. A new zero-day may have no score yet; that is missing data, not low risk.

What should we do in the first hour after a zero-day is announced?

Open an event with a named commander, preserve logs for the affected product, and find every instance, internet-facing and management interfaces first. Apply vendor mitigations or restrict access on exposed systems, hunt from the earliest plausible exploitation date, and hand any asset with indicators of exploitation to incident response.

Can a penetration test prove we were not compromised?

No. Penetration testing can verify exposure, confirm whether an instance is exploitable as configured, test whether a mitigation blocks the vulnerable path, and show what an attacker could reach. Establishing whether compromise occurred is the work of incident response and digital forensics, using logs, telemetry and host evidence.

Is it safe to test a zero-day exploit against production systems?

No. Untrusted public exploit code should never be run against production. Authorized validation uses version and configuration checks, vendor tools, non-intrusive mitigation checks and, where proof is needed, controlled testing in a production-equivalent replica, always with written authorization, rules of engagement, and a pause on assets under incident-response control.

When does a zero-day become an incident-response matter?

As soon as an asset shows indicators of exploitation, it moves to incident-response command, and evidence is preserved before patching or rebuilding unless the incident lead decides otherwise. Many organizations involve incident response earlier when exploitation is widespread and their telemetry on the product is weak, because waiting for certainty can cost evidence.

Do zero-day exposures trigger breach notification in Canada or the US?

Exposure alone usually does not; confirmed compromise can. In Canada, PIPEDA requires reporting breaches that create a real risk of significant harm, and OSFI expects qualifying incidents reported within 24 hours. In the US, SEC rules cover material incidents and sector rules such as NYDFS Part 500 apply. Counsel decides, using your triage record.

Conclusion

A zero-day vulnerability does not have to become an incident, but uncertainty left unmanaged often lets it become one. What keeps it from becoming one is the discipline described here: a shared status vocabulary, verified exposure, verified mitigation, and a clean line between exposure and compromise. Start with the free playbook, and if you need independent evidence on whether a specific exposure is real and what an attacker could reach from it, talk to us about authorized validation.

Book a Zero-Day Exposure Scoping Call. A short conversation to scope authorized exposure, exploitability and attack-path validation for a specific vulnerability. Book a Zero-Day Exposure Scoping Call. Booking a call does not start an emergency or incident-response engagement; if you suspect active compromise, contact your incident-response provider first.

Need this tested?

Our Network Penetration Testing service validates perimeter exposure, exploitability and attack paths under written authorization. Book a Zero-Day Exposure Scoping Call to discuss scope and timing. Booking a call does not start an emergency or incident-response engagement.

RELATED ARTICLES
Explore Network Penetration Testing →