---
service: "Publicasta"
schema_version: "1.0"
article_id: 1211
title: "CVE-2026-124194 in Outlook: Why an Exploited NTLM Hash Leak Needs More Than an Office Update"
language: "en"
default_language: "en"
canonical_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=en"
json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=en"
api_url: "https://publicasta.com/api/public/v1/channels/cybersecurity/articles/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=en"
channel_url: "https://publicasta.com/api/public/v1/channels/cybersecurity"
channel_articles: "https://publicasta.com/api/public/v1/channels/cybersecurity/articles"
search_url: "https://publicasta.com/api/public/v1/search"
documentation_url: "https://publicasta.com/api-docs#reading-publicasta"
openapi_url: "https://publicasta.com/api-docs/openapi.json"
published_at: "2026-11-12T13:46:21+00:00"
updated_at: "2026-11-12T13:46:21+00:00"
translations:
  - language: "ar"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=ar"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=ar"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=ar"
  - language: "de"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=de"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=de"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=de"
  - language: "en"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=en"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=en"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=en"
  - language: "es"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=es"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=es"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=es"
  - language: "fr"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=fr"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=fr"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=fr"
  - language: "pl"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=pl"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=pl"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=pl"
  - language: "ru"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=ru"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=ru"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=ru"
  - language: "zh"
    html_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response?lang=zh"
    markdown_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.md?lang=zh"
    json_url: "https://publicasta.com/cybersecurity/outlook_cve_2026_124194_ntlm_hash_leak_response.json?lang=zh"
---

# CVE-2026-124194 in Outlook: Why an Exploited NTLM Hash Leak Needs More Than an Office Update

> Microsoft has patched an exploited Outlook flaw that can disclose a user’s NTLMv2 authentication response when a crafted message file is opened. The immediate task is to update Office, then reduce the network paths that make leaked authentication material useful.

Microsoft’s November 10, 2026 security release includes a vulnerability in Outlook that attackers were already exploiting before a fix became available. Tracked as **CVE-2026-124194**, the flaw can cause Outlook to connect to an attacker-controlled network location when a specially crafted message file or Outlook item is opened. That connection can disclose the user’s NTLMv2 authentication response.

 ![Security workstation displaying a suspicious email attachment and an outbound authentication connection to a remote server](https://publicasta.com/storage/projects/9/pages/1211/2026/11/076a5d2e-dce6-4814-9ed4-f31f4454b174.webp)

 This is not a remote-code-execution bug, and Microsoft rates it as Moderate with a CVSS score of 6.5. Those numbers can make the issue look less urgent than a critical server vulnerability. The exploitation status matters more here: Microsoft says exploitation has been detected, and the vulnerability has been added to the U.S. Cybersecurity and Infrastructure Security Agency’s Known Exploited Vulnerabilities catalog.

 The practical response has two parts. First, install the November Office security update across every affected Outlook installation. Second, treat any endpoint that may have opened an untrusted message file before patching as a credential-exposure case, and review the organization’s controls for outbound NTLM and SMB traffic.

 ## What the vulnerability changes

 CVE-2026-124194 is a spoofing and forced-authentication issue in Microsoft Office Outlook. The problem is not that Outlook simply displays malicious content. The problem is that a crafted file can make the application initiate an external network connection without giving the user the warning that should normally accompany that action.

 In a typical Windows environment, a connection to a network resource can cause the operating system to attempt integrated authentication. With NTLM, the client does not send the user’s plaintext password, but it can send a challenge-response value derived from the account’s secret. That value is commonly called an NTLMv2 or Net-NTLMv2 response. An attacker who controls the destination can collect it.

 A captured response is not the same thing as a password database dump. It is also not automatically equivalent to unrestricted access to the account. But it is authentication material, and its value depends on the rest of the environment. It may be used in relay attacks where the captured response is forwarded to a service that accepts NTLM, or it may be subjected to offline password guessing. The risk increases when the affected user has broad access, uses a weak or reused password, or can authenticate to internal services that do not require stronger protections.

 The vulnerability therefore sits at an uncomfortable boundary. The visible event is a user opening a file. The security consequence is a network authentication attempt that can cross the organization’s perimeter. A mail-client patch closes the specific Outlook behavior, but network and identity controls determine how much damage a leaked response can do.

 ## Who is affected

 Microsoft’s November Office release notes list CVE-2026-124194 under Outlook. The affected product families include:

 - Microsoft 365 Apps for enterprise before the November 10, 2026 security update.
- Office LTSC 2024 before build **16.0.17932.21140**.
- Office LTSC 2021 before build **16.0.14334.21048**.

 Microsoft’s release notes also identify the current Microsoft 365 Apps builds published on November 10, including version 2610 builds for the Current Channel and Monthly Enterprise Channel. The exact build that applies depends on the update channel and deployment model, so administrators should use Microsoft’s Office security release notes and their normal update-management inventory rather than rely on a single version number.

 The scope is narrower than “every Microsoft email user.” Exchange Online service-side protections do not make every Outlook client irrelevant, and organizations often have a mixture of Microsoft 365 Apps, perpetual Office, virtual desktops, shared workstations, and older installations. The important question is not whether a company has Microsoft 365 somewhere in its environment. It is whether a Windows endpoint runs an affected Office build and can be induced to open a crafted Outlook-related file.

 Administrators should inventory both managed and less visible installations. That includes terminal servers, jump hosts, executive laptops, contractor devices covered by endpoint management, and machines used for mailbox support or e-discovery. A vulnerable Outlook installation that is rarely used can still matter if it contains a privileged user profile or has access to internal shares.

 ## Why the CVSS score is not the priority signal

 A CVSS score describes technical characteristics of a vulnerability. It does not describe how quickly attackers are using it in the real world, how many vulnerable systems an organization owns, or how exposed those systems are.

 CVE-2026-124194 requires user interaction: the crafted file has to be opened. That requirement lowers its CVSS score compared with a no-click network service flaw. The impact category also emphasizes confidentiality rather than immediate code execution or availability loss. Those facts explain the 6.5 score, but they do not cancel the exploitation warning.

 The right prioritization question is: can an attacker cause this behavior on a machine whose user can authenticate to valuable systems? In many organizations, the answer is yes. Email remains a reliable way to deliver files, and Outlook users may open files received through shared mailboxes, ticketing systems, external collaboration, or forwarded messages. A vulnerability that converts one opening action into an outbound authentication attempt deserves faster treatment than an unexploited flaw with a higher theoretical score.

 CISA’s KEV catalog is useful for this reason. It is not a complete measure of risk, but an entry signals that defenders should treat exploitation as an operational fact. Federal civilian agencies receive a remediation deadline under the catalog process; other organizations should not interpret that date as a safe waiting period.

 ## What to do first

 ### 1. Confirm the Office update path

 Apply Microsoft’s November 2026 Office security update through the organization’s normal servicing channel. For Microsoft 365 Apps, check the update channel and build after deployment. For Office LTSC, verify that the installed build is at or above the fixed version. Do not close the ticket merely because an update was approved in a management console; confirm that representative endpoints actually received it and restarted the relevant Office applications.

 Prioritize machines used by administrators, finance teams, help-desk staff, executives, and anyone with access to sensitive file shares or management systems. This is an exposure-based ordering choice, not a claim that other users are safe. After the first urgent wave, complete the rollout across the full estate.

 If an endpoint is offline, rarely connected, or outside the organization’s normal management boundary, record it explicitly. A missing device in the inventory should be treated as an unresolved patch state rather than as a device that probably does not matter.

 ### 2. Separate patching from investigation

 Patching stops the known Outlook trigger. It does not prove that no one opened a malicious file before the update.

 Ask the service desk, email-security team, and incident-response team whether users reported unusual message files, unexpected authentication prompts, or suspicious external mail during the period before patching. Preserve relevant mailbox and endpoint telemetry according to the organization’s retention policy. Look for outbound authentication attempts from user workstations to unfamiliar systems, especially over SMB or other protocols that should not leave the trusted network.

 Avoid turning every unusual network event into a breach declaration. A connection may be a false positive, a legitimate business workflow, or a different Outlook feature. The objective is to identify endpoints and identities that need closer review, not to infer an attacker from one log line.

 If there is credible evidence that an endpoint opened a malicious file and exposed authentication material, isolate the machine from sensitive network segments while the response team investigates. Review the user’s recent sign-ins, access to internal shares, authentication failures, and unusual activity involving privileged services. Password changes may be appropriate after exposure, but they should be coordinated with the identity team and accompanied by token or session review where the organization’s identity platform supports it.

 ### 3. Reduce outbound NTLM exposure

 Microsoft’s guidance for restricting outgoing NTLM recommends starting with auditing. The Windows policy **Network security: Restrict NTLM: Outgoing NTLM traffic to remote servers** can audit outbound NTLM attempts before administrators move toward denying them. Audit data helps identify legitimate dependencies and exceptions.

 For newer Windows environments, Microsoft also documents NTLM blocking capabilities for SMB clients. These controls can prevent clients from sending NTLM authentication to remote servers, but they can affect legacy applications, appliances, and file-sharing workflows. A broad deny rule without an inventory of dependencies may create outages and encourage emergency exceptions that are harder to govern.

 A safer sequence is to measure, document, reduce, and then deny where the business case allows it:

 1. Enable auditing and collect the operational events generated by outbound NTLM policy.
2. Identify which applications and destinations still depend on NTLM.
3. Move supported services to Kerberos or another stronger authentication method.
4. Create tightly scoped exceptions only where there is a documented need.
5. Enforce blocking for destinations and device groups that should never receive NTLM.

 Do not treat NTLM restrictions as a substitute for the Outlook update. They are a second control that limits the usefulness of a leaked authentication response and addresses other forced-authentication paths as well.

 ### 4. Block outbound SMB at the perimeter

 Microsoft recommends blocking outbound TCP port 445 to the internet. Most organizations do not need ordinary Windows workstations to establish SMB sessions with arbitrary external addresses. Blocking that traffic reduces the chance that an Outlook-triggered connection can send authentication material directly to an internet host.

 This control should be implemented at corporate firewalls and on endpoint networks where users work remotely. A perimeter rule alone may not protect a laptop on a home or public network, and a local rule alone may not cover every egress path. Review IPv6, VPN split tunneling, cloud security gateways, and alternative file-sharing protocols as part of the network design.

 Port blocking is not a complete answer. A forced-authentication attempt might use a protocol other than SMB, and an internal attacker-controlled system may be reachable even when the internet is not. The goal is to remove unnecessary external paths while identity and endpoint controls address the remaining cases.

 ### 5. Require appropriate SMB protections

 SMB signing helps prevent tampering and relay attacks by allowing the client and server to verify the integrity and identity context of SMB traffic. Microsoft says Windows 11 version 24H2 and Windows Server 2025 require SMB signing by default in more configurations than earlier releases, but organizations should verify the effective policy rather than assume the default applies everywhere.

 Signing does not make a stolen NTLM response harmless, and it is not the same as blocking NTLM. It is one layer in a broader control set. Where compatible, use Kerberos instead of NTLM, require signing, and use encryption where the deployment needs confidentiality as well as integrity. Test older NAS devices, printers, application servers, and third-party file services before changing requirements globally.

 The migration work is worthwhile beyond this single CVE. Outbound NTLM and unsigned SMB create exposure to a family of credential-relay and forced-authentication problems. A documented exception that remains visible is safer than an undocumented legacy dependency that nobody remembers until the next incident.

 ## What not to overclaim

 There are several common ways to describe this vulnerability inaccurately.

 First, a leaked NTLMv2 response is not the same as the plaintext password. The response can still be dangerous, especially in relay scenarios, but the technical distinction matters when deciding whether to reset passwords, invalidate sessions, investigate lateral movement, or notify affected parties.

 Second, the issue should not automatically be described as a full Outlook remote-code-execution vulnerability. The reported impact is disclosure of authentication material through an unwanted external connection. That is serious, but the response should match the actual behavior.

 Third, a patch deployment does not prove that an earlier exposure did not occur. Teams should retain enough endpoint, email, proxy, firewall, and identity telemetry to investigate the period before remediation. Conversely, the presence of a suspicious message does not prove that the recipient opened it or that authentication material was successfully collected.

 Finally, public discussion may provide useful context, but it should not be used to invent a threat actor, victim list, or campaign scope that Microsoft and CISA have not published. Defenders can act decisively on the confirmed facts: the flaw was exploited, a fix is available, affected Office builds exist, and NTLM controls can reduce downstream risk.

 ## A practical validation checklist

 After remediation, a security or desktop-engineering team should be able to answer these questions with evidence:

 - Which devices run Microsoft 365 Apps, Office LTSC 2024, or Office LTSC 2021?
- Which devices were below the fixed Office build on November 10, 2026?
- Which devices have now installed and activated the update?
- Which users or endpoints opened untrusted `.msg` or Outlook-related files during the exposure window?
- Did any workstation attempt outbound NTLM authentication to an unusual external or internal destination?
- Is TCP 445 blocked outbound from user networks, including remote-access paths?
- Is outbound NTLM audited, restricted, or blocked, and are exceptions documented?
- Are SMB signing and Kerberos available for the services that still handle sensitive files?
- Have identity and incident-response teams reviewed accounts associated with credible exposure?
- Can the organization demonstrate that update compliance includes offline and unmanaged devices?

 The quality of the response is visible in the evidence, not in the number of emergency messages sent to staff. A good closure record shows the affected population, the fixed build, the deployment date, the remaining exceptions, and the compensating controls for systems that could not be updated immediately.

 ## The broader lesson for vulnerability management

 CVE-2026-124194 is a useful example of why vulnerability response cannot be reduced to a list sorted by CVSS. The flaw is in a common desktop application, requires a user action, and produces a credential leak rather than immediate code execution. In isolation, those properties may suggest a medium-priority ticket. Add confirmed exploitation, Windows integrated authentication, access to internal services, and the possibility of outbound network reachability, and the operational priority changes.

 The most resilient organizations make this kind of decision quickly because they already know where Office is installed, which endpoints can reach sensitive services, where NTLM remains in use, and which logs can show outbound authentication. They do not need to discover their inventory during the incident.

 For everyone else, the immediate course is still clear: deploy the November Outlook update, verify the result, inspect the pre-patch exposure window, and reduce the network and identity paths that allow a leaked authentication response to become an account compromise. That work addresses this exploited vulnerability while improving defenses against the next forced-authentication flaw.
