On this pageOrigin of the TermWhat Counts as an ExposureAttack Paths and Choke PointsStages of a CTEM ProgramOwnership Across the StagesExisting Practices Mapped to StagesCTEM and Vulnerability Management ComparedValidation Results and Scan ScoresStarting With One ScopeOne Cycle Followed ThroughRecords Kept at Each StageHow a CTEM Program Is MeasuredCTEM as a Managed ServiceExposure Management With ThreatDefenceFrequently Asked Questions Origin of the TermThe term was introduced by industry analysts in 2022. It described a problem most security teams already recognized: scanners and assessments produce more findings than a team can fix, and a severity score alone says little about which of them an attacker could use. The proposal was to treat exposure as a continuous program with a defined scope and a business owner, in place of a list of vulnerabilities worked from the top down. Each cycle ends with changes made and verified, and the next cycle starts from what was learned. The idea has since appeared in product design and in public policy, and the pressures behind it have grown since 2022. Product design: Microsoft’s documentation describes its Security Exposure Management product as aligned with the CTEM approach. Public policy: In June 2026 CISA issued Binding Operational Directive 26-04, which tells US federal civilian agencies to prioritize security updates by risk. Breadth: Many routes into an organization involve a misconfiguration, a leaked password or an unknown host, where no patch applies. Pace: The same CISA directive observes that attackers’ use of AI may further shorten the time defenders have to react, a change our guide to AI hacking covers. What Counts as an ExposureAn exposure is any condition an attacker could use to gain access or to move closer to something valuable. A software vulnerability with a CVE identifier is one kind among several. Unpatched software: A product version is affected by a published vulnerability, above all one listed in CISA’s Known Exploited Vulnerabilities (KEV) catalog. Misconfiguration: A storage bucket is open to the public, a management console is reachable from the internet, or a default password is left in place. Identity weakness: Accounts lack multi-factor authentication, privileged accounts go stale, or applications hold more permissions than they use. Leaked credentials: Valid usernames and passwords for corporate accounts appear in breach data or infostealer logs. Unknown assets: Hosts, subdomains and cloud resources sit outside the inventory, so they are typically left unpatched and unmonitored. Detection gaps: An attack technique would succeed without raising an alert. Exposures matter most in combination. A forgotten test server, a reused password and an over-privileged account may each look minor, and together they can form a route to production data. Microsoft’s exposure management documentation calls these routes attack paths. Attack Paths and Choke PointsAn attack path is the chain of assets and techniques that leads from a point an attacker can reach to an asset the organization treats as critical. CISA’s guide to Stakeholder-Specific Vulnerability Categorization (SSVC) calls the same idea vulnerability chaining, the use of several weaknesses to reach one outcome. Each exposure is judged by the part it plays in a path. Entry point: The path starts on the attack surface, at something an outsider can reach, such as an internet-facing host or a leaked credential. Intermediate hop: Each hop is a weakness that moves the attacker on, for example a stored password or a group membership. Target: The path ends at a critical asset named during scoping, such as a domain controller, a database holding sensitive data or a privileged role. Choke point: Microsoft uses this name for an asset where several paths converge on the way to a target. Thinking in paths changes the order of work. Position outranks score: A medium-severity flaw on an entry point may deserve attention before a critical flaw on a host that leads nowhere. Choke points come first: Securing one asset that many paths cross closes all of them. One barrier can break a path: The SSVC guide notes that a single effective barrier, such as a requirement to be signed in, can keep exploitation from being automated, so a path may be cut at its cheapest hop. Paths are recomputed: Microsoft notes that the set of paths shifts as assets are added, configurations change and users join groups. Path analysis depends on scoping, and Microsoft’s documentation warns that attack paths may be less than fully representative where critical assets have yet to be defined. A path drawn from inventory data also stays an estimate until validation tests each hop. Stages of a CTEM ProgramThe stages run in a fixed order, and the output of each one is the input of the next. Scoping Scoping decides which part of the organization the cycle covers and what an attacker would want from it. A scope is named in business terms, such as the customer portal, the payment environment or the Microsoft 365 tenant, and it has an owner who can approve changes. Goes in: Scoping starts from business priorities, the systems that support them, recent incidents and the concerns of system owners. Comes out: The stage produces a written scope with its owner, the critical assets inside it, and the outcomes the business treats as unacceptable. Discovery Discovery lists the assets in scope and the exposures on them. It reaches past a scan of known addresses to hostnames and certificates, cloud resources, identities and their privileges, configuration, and credentials that have already leaked. Goes in: Discovery takes the scope, seed domains and address ranges, read access to cloud and identity tenants, and scanner results. Comes out: The stage produces an asset inventory and a raw list of exposures, each tied to an asset and an owner. Discovery of internet-facing assets is explained source by source in our external attack surface management guide, and vulnerability scanning supplies the software-flaw part of the list. Prioritization Prioritization orders the exposures by how likely an attacker is to use each one and what it would give them. A severity score such as CVSS describes the flaw. Priority also depends on the asset the flaw sits on. BOD 26-04 is a public model for this. It asks agencies to weigh the following for each vulnerability, and CISA encourages organizations outside the federal government to adopt similar practice. Asset exposure: The first question is whether the vulnerable asset is publicly exposed. Known exploitation: The next is whether the vulnerability appears in the KEV catalog, which CISA says organizations should use as an input to prioritization. Exploit automation: The directive then asks whether an adversary could automate all the steps of exploitation. Technical impact: The last question is whether a successful exploit gives partial or total control of the asset. A CTEM program applies the same reasoning to exposures that have no CVE, such as a leaked administrator password for an internet-facing sign-in page. Goes in: Prioritization takes the raw exposure list, asset criticality from scoping, and exploitation intelligence. Comes out: The stage produces a short ordered list for validation, and a longer list handed to routine patching. Validation Validation tests whether a prioritized exposure can be used, how far it leads, and whether existing controls stop or detect the attempt. An exposure behind a control that holds typically drops in priority, and one that chains into privileged access rises. Goes in: Validation needs the ordered list, written authorization to test, and rules for safe testing. Comes out: The stage produces confirmed exposures with proof, exposures shown to be blocked, and a note of what monitoring saw. Validation is where penetration testing joins the program. Because the list changes every cycle, many programs add AI penetration testing, in which agents test new exposures as they are found. Mobilization Mobilization gets confirmed exposures fixed by the teams that own the systems. It typically takes the most coordination, because the security team rarely owns the fix. Goes in: Mobilization receives confirmed exposures with proof, fix options and a named owner for each. Comes out: Each exposure ends as a completed change with a passing retest, or as a risk acceptance recorded by the owner with a review date. Mobilization also asks whether the exposure was used before it was closed. CISA’s announcement of BOD 26-04 makes the point that applying a patch does not generally remove an attacker who is already inside. The directive pairs its three-day remediation timeline with a forensic triage of the asset. Ownership Across the StagesEach stage has its own owner, and the security team leads only some of them. A program lead, typically the security lead, keeps the queue and reports the measures. Stage Who leads Who contributes Decision held Scoping Business or system owner Security lead, risk function What is in scope and which outcomes are unacceptable. Discovery Security or IT operations Cloud, identity and network administrators Whether a found asset belongs to the organization, and who owns it. Prioritization Security lead Threat intelligence, asset owners Which exposures go to validation and which go to routine patching. Validation Internal tester or testing provider The system owner who authorizes the test, the SOC Whether an exposure is confirmed, blocked or left untested. Mobilization System owner IT operations, change management, security lead Whether to fix now, mitigate, or accept the risk until a review date. Public frameworks expect the roles to be written down. BOD 26-04 requires agency vulnerability management policies to assign roles and responsibilities and to set internal validation, tracking and reporting requirements. The NCSC guidance describes a decision to leave an update unapplied as a senior-level risk decision, which is why acceptance sits with the owner of the scope. In a small organization one IT manager may hold scoping, prioritization and mobilization, with discovery and validation supplied from outside. In a large one each scope has its own business owner, and a central team keeps the method consistent between scopes. Existing Practices Mapped to StagesMost organizations already run several of the activities a CTEM program needs. The program connects them into one queue. Existing practice Stage it serves What CTEM adds Asset inventory Scoping, discovery Criticality for each scope, and the assets the inventory missed. External attack surface management Discovery An owner and a scope for every internet-facing asset found. Vulnerability scanning Discovery Context on exposure and exploitation, so results are ordered by more than severity. Cloud and identity posture review Discovery Misconfigurations and privileges joined to software flaws as attack paths. Threat intelligence and the KEV catalog Prioritization A match against the organization’s own exposed products. Penetration testing Validation Tests chosen from the priority list and repeated when the scope changes. Red and purple teaming Validation Detection and response results fed into the same queue as technical findings. Patch and change management Mobilization Proof attached to each request, and a retest before closure. Our guide to penetration testing explains the method validation relies on, and its related guides compare it with red teaming and with vulnerability scanning. CTEM and Vulnerability Management ComparedVulnerability management finds and patches software flaws across the estate, and CTEM is a wider program that uses its results. They differ in what is counted, how it is ordered and what closes an item. Vulnerability management CTEM Unit of work A vulnerability on an asset. An exposure inside a business scope. Covers Software flaws, mostly those with a CVE. Software flaws plus misconfiguration, identity weakness, leaked credentials, unknown assets and detection gaps. Ordering Severity score, sometimes with exploit data. Asset exposure, known exploitation, and what an attacker would reach. Evidence A scanner detection. A test result with proof. Closure Patch applied or scan clean. Fix confirmed by retest, or risk accepted by the owner. Reporting Open vulnerabilities by severity. Confirmed exposures closed, and time taken to close them. NIST SP 800-40 Rev. 4 describes patching as preventive maintenance for technology, and that routine work continues on its normal schedule inside a CTEM program. CTEM decides which items leave the routine queue for testing and urgent action. It also covers the cases a patch cannot settle. The UK NCSC’s vulnerability management guidance states the principle that the organization must own the risks of not updating, and mobilization is where that ownership is recorded. Validation Results and Scan ScoresA scan score rates a flaw in general, and a validation result records what happened when someone tried to use it in one environment. A score such as CVSS is assigned to the vulnerability once, for every organization that has it, as our vulnerability scanning guide explains. A validation result belongs to one asset on one date. A test settles points that a score leaves open: whether the weakness can be reached, whether it works against this configuration, how far it leads, and whether a control stopped or detected the attempt. Each exposure then leaves validation in a recorded state, and the state decides its place in the queue. Confirmed: The attempt worked and proof was kept, so the item moves to mobilization with its tested priority. Confirmed and undetected: The attempt worked and monitoring stayed silent, so a second item is raised for the detection gap. Blocked: A control stopped the attempt, so priority drops and the control is recorded as one the scope depends on. Unreachable: The weakness exists but cannot be reached from the tested position, so it returns to routine patching. Untested: Authorization, safety rules or time ruled the test out, so the item keeps its estimated priority. Automated checks can settle part of this. Microsoft describes active reachability scans that validate whether a cloud exposure is accessible from outside the environment, and the remaining points typically need an authorized test. Starting With One ScopeA first CTEM cycle works best on a single scope that is small enough to finish. The internet-facing estate is a common first choice, because discovery needs only the organization’s domains. A first quarter is typically enough to take that scope through a full cycle and to leave a working routine behind. Weeks 1 and 2, agreement: The program lead and the business owner sign the scope statement, list the critical assets and settle who authorizes testing. Weeks 3 and 4, first discovery: Discovery runs across the scope, and differences from the inventory are reviewed with the administrators who know the assets. An asset that nobody claims gets an interim owner. Weeks 5 and 6, ordering: The exposure list is ordered, likely paths to the critical assets are sketched, and the validation list is cut to what can be tested. Weeks 7 to 9, validation: Tests run inside agreed windows. Confirmed items are handed over as they are found. Weeks 10 to 12, fixes and retests: Owners work the items and retests close them. Items that cannot be fixed in time are mitigated or accepted on the record. End of quarter, review: The lead reports the measures, the owner signs any accepted risks, and both agree what changes in the second cycle. The quarter is also the time to set working habits that later scopes inherit. A weekly queue review: The lead and the system owners go through what was confirmed, what is overdue and what is stuck. A fast lane: A written rule states which events bypass the cycle, such as a KEV addition for an exposed product. Later cycles add scopes one at a time, typically identity, then cloud, then internal networks. Between cycles, the events described in our guide to continuous penetration testing start a test of their own. One Cycle Followed ThroughOne cycle on a customer portal scope at a mid-sized company shows how each stage changes what the team knows. Scoping. The head of customer operations owns the scope, the critical asset is the customer database, and the unacceptable outcome is customer records read by an outsider. Discovery. Certificate records show a staging copy of the portal that is absent from the inventory. The scanner reports a medium-severity flaw on it, and a credential source lists a leaked password for a contractor’s account. Prioritization. Each finding alone would wait for routine work. Together they suggest a path from a public host through a valid account to the production database, which goes to the top of the validation list. Validation. With written authorization, the tester confirms that the leaked password still signs in to the staging portal and that the portal returns live customer data. The tester stops at a single record and keeps the request and the response. Monitoring raised no alert for the sign-in. Mobilization. The identity administrator resets the account and enforces multi-factor authentication, and the platform team takes the staging host off the internet. The SOC adds a detection for sign-ins to non-production hosts and searches earlier logs for use of the leaked password. Retest and review. The tester repeats the sign-in and the read, both fail, and the items close. The review adds non-production hostnames to the scope statement. On severity alone, none of these findings would have reached the top of a patching queue. The cycle also produced an inventory entry, a new detection, and a log search that answered whether the route had already been used. Records Kept at Each StageEach stage of a CTEM program leaves a record, and those records are how the program is audited and improved. Record Stage Where it comes from Scope statement and critical asset list Scoping Agreed with the business owner and stored with the risk register. Hostname and certificate inventory Discovery DNS zone exports and Certificate Transparency logs for the organization’s domains. Cloud resource and privileged role inventory Discovery The cloud provider’s asset inventory and the identity directory’s role assignment export. Vulnerability results per asset Discovery Scanner exports, kept with the scan date. KEV match list Prioritization CISA’s KEV catalog, published as CSV and JSON, joined to the software inventory. Priority decision Prioritization The recorded outcome for each exposure, for example CISA’s SSVC decisions of Track, Track*, Attend or Act. Proof of exploitability Validation The tester’s request and response, or command output, with the time and source address. Detection check Validation The alert, or the absence of one, in the monitoring platform for the test window. Fix and retest result Mobilization The change ticket and the retest record that closed it. Risk acceptance Mobilization A risk register entry with the owner’s name and a review date. These records line up with outcomes in the NIST Cybersecurity Framework 2.0. ID.AM-05 asks that assets are prioritized by classification, criticality, resources and impact on the mission. ID.RA-01 asks that vulnerabilities in assets are identified, validated and recorded, and ID.RA-06 that risk responses are chosen, prioritized, planned, tracked and communicated. How a CTEM Program Is MeasuredA CTEM program is measured by what was confirmed and closed in each scope. Counts of open vulnerabilities describe the size of the backlog, and say little about whether an attacker’s routes are being removed. Discovery coverage: The measure is the share of in-scope assets seen in the latest cycle, with the number found outside the inventory. Time to inventory: The clock runs from the day a new internet-facing asset appears to the day discovery records it. Validation rate: The rate is the share of high-priority exposures that were tested, and the share of those confirmed. Time to fix: The program counts the days from confirmation to a passing retest, reported for each owner. Retest results: The figure is the share of reported fixes that hold when tested again. Overdue acceptances: The count covers accepted risks that have passed their review date. Detection during validation: The measure is the share of validation tests that raised an alert in monitoring. Trends per scope are more useful than a single figure. A falling time to fix with a steady validation rate suggests mobilization is working, and a rising count of unknown assets points back to discovery. CTEM as a Managed ServiceA managed service can run most of the cycle for an organization, while decisions about scope and risk stay inside it. Discovery by the provider: The provider maps the scope continuously and reports what changed, and the organization confirms which assets are its own. Validation by the provider: The provider orders the list by the agreed rules, tests the top of it inside the written authorization, and keeps the proof. Tracking by the provider: The provider keeps the queue current and retests each reported fix. Scope and authorization from the organization: The business owner names the scope and its critical assets, and the system owner signs the permission to test. Fixes and acceptances by the organization: System owners make the changes, and only the organization can accept a risk. A public service works on the same division. CISA’s Cyber Hygiene services scan the internet-accessible assets of enrolled organizations and report findings weekly, and BOD 26-04 has federal agencies update or attest their publicly exposed IP addresses and agency-owned domain names once per quarter. Some points are worth settling before a provider starts. Stages covered: The agreement states which stages the service runs. Link to monitoring: The testers and the monitoring team share results, so a detection check and a search for earlier use follow each confirmed exposure. Records on exit: The inventory, the proof and the acceptance records remain the organization’s if the service ends. An MSP or MSSP can run the same service for its clients, with one scope and one authorization per client. Our guide to white label penetration testing describes that model for the validation stage. Exposure Management With ThreatDefenceThreatDefence supports the discovery, validation and mobilization stages of a CTEM program for the internet-facing scope. The agents belong to its AI SOC platform, which automates monitoring, compliance and cyber risk management and has a 24x7 SOC behind it. Discovery: The ASM agent maps the domains, hosts, services and certificates an organization exposes and notes each change, and the Brand agent adds leaked credentials and lookalike domains. Validation: In agentic penetration testing, the testing agent tests those systems inside an authorized scope and keeps proof for each weakness it finds exploitable, with SOC analysts reviewing significant findings. Mobilization: Every problem becomes one tracked action item, and the testing agent retests once the owner reports the fix. Scoping stays with the organization, which confirms the scope before any test runs. Where a customer is monitored, the Defender and Hunter agents also look through its telemetry for earlier use of a confirmed weakness. Internal networks and social engineering fall outside the agents’ testing, and an independent human-led test can cover them alongside. Frequently Asked QuestionsWhat does CTEM stand for?CTEM stands for continuous threat exposure management. It names a repeating program in which an organization scopes part of its environment, discovers the exposures in it, orders them by risk, tests which can be used, and gets the confirmed ones fixed. Industry analysts introduced the term in 2022, and it is now used by practitioners and product makers alike.Is CTEM a product?No. CTEM is a program made of people, process and several tools, and no single product delivers every stage. Discovery, scanning, testing and ticketing tools each serve one part. The program supplies the scope, the ordering rules, the handover to system owners and the retest, which are the parts a tool cannot decide on its own.Does CTEM replace vulnerability management?No. Vulnerability management continues as the routine process that finds and patches software flaws, and CTEM draws on its results. CTEM adds exposures a scanner does not report, such as leaked credentials and misconfigured identity, orders everything by exposure and exploitation, and requires proof before urgent work is asked of a system owner.What are the stages of CTEM?The stages are scoping, discovery, prioritization, validation and mobilization. Scoping sets the boundary and the owner. Discovery lists assets and exposures. Prioritization orders them. Validation tests which can be used. Mobilization gets them fixed and retested. The cycle then repeats, with later cycles widening the scope or returning to one that has changed.How is CTEM different from exposure management?Exposure management is the general discipline of knowing and reducing the ways an organization can be attacked. CTEM is one specific model for running it, defined by a repeating cycle with named stages and a validation step. An organization may practice exposure management without using the CTEM stages, though most programs end up with similar steps.Is CTEM required by a standard or regulation?No. CTEM is an analyst model, and the standards named in this guide do not require it by name. Its activities do map to recognized outcomes. The NIST Cybersecurity Framework 2.0 expects asset inventories, validated vulnerabilities and tracked risk responses, and CISA’s BOD 26-04 requires US federal civilian agencies to prioritize security updates by risk.Can a small security team run CTEM?Yes. A small team can run one scope at a time, typically the internet-facing estate, using its existing scanner, a discovery source and a ticket queue. The stage that most often needs outside help is validation, which a testing provider can supply. The discipline that matters is finishing each cycle with fixes retested before adding a new scope.Where does a SOC fit in a CTEM program?A SOC contributes to validation and mobilization. During validation it shows whether a test raised an alert, which turns a detection gap into a finding of its own. During mobilization it searches logs for signs that a confirmed exposure was already used, and it can watch the affected system more closely until the fix is in place.What is an attack path in exposure management?An attack path is a chain of exposures that leads from a point an attacker can reach to an asset the organization treats as critical. CTEM uses attack paths to order work. An exposure on a path to a critical asset is tested and fixed ahead of a higher-scored flaw that leads nowhere, and an asset where several paths meet is fixed first.How long does a CTEM cycle take?The length varies with the scope. A first cycle on one scope typically fits within a quarter, because the scope, the owners and the authorization to test are being agreed for the first time. Later cycles on the same scope run faster, and events such as a newly exposed host are handled between cycles as they occur.