Automated Threat Modeling

Quick Look: FortiMail CVE-2026-126151 Authentication Flaw Enabling Unauthenticated Account Takeover

A brief summary of CVE-2026-126151, an incorrectly implemented authentication algorithm (CWE 303) in Fortinet FortiMail that may let an unauthenticated remote attacker take over an account via a crafted request. Covers affected branches, fixed firmware builds, and advisory inconsistencies.

CVE Analysis

7 min read

Gaurav Sarraf
Gaurav Sarraf

2026-11-11

Quick Look: FortiMail CVE-2026-126151 Authentication Flaw Enabling Unauthenticated Account Takeover

Introduction

Fortinet has fixed a flaw in FortiMail's authentication implementation that, according to NVD, may allow an unauthenticated remote attacker to take over an account using a single crafted request. FortiMail sits directly in the mail path of the organizations that deploy it, so an authentication failure reachable over the network without credentials deserves attention, even though Fortinet currently reports no known exploitation.

FortiMail is Fortinet's email security product. It draws on FortiGuard services, including antivirus, virus outbreak protection and antispam, and Fortinet offers it in several deployment models, including FortiMail Cloud, which delivers email security as a service. Fortinet reports a portfolio of over 50 enterprise-grade products and more than 900,000 customers. These are companywide figures, not a count of FortiMail installations. That footprint, combined with a separate FortiMail flaw reported under active exploitation in October 2026, makes it worth reading this advisory carefully rather than lumping it in with other recent FortiMail news.

Technical Information

Root cause: CWE-303

The weakness is classified as CWE-303, Incorrect Implementation of Authentication Algorithm. MITRE defines this as a product that requires an established authentication algorithm but implements it incorrectly, which may allow authentication to be bypassed. The authentication scheme itself exists. The problem is in how it is carried out.

Fortinet's advisory FG-IR-26-203 labels the affected component as "API" and the attack type as "Unauthenticated". NVD describes the outcome as elevation of privilege in the form of account takeover, triggered by a crafted request from an unauthenticated remote attacker.

That is where the public mechanism stops. The advisory does not identify:

  • the request path or HTTP method
  • the parameters involved
  • the authentication exchange that is mishandled
  • the type of account that can be taken over
  • the privilege level the attacker ends up with

FortiMail is closed source, so there is no public commit, diff, or vulnerable code to examine, and Fortinet has not published root cause details. Going by the CWE label, the defect is likely something like a broken comparison, a token or signature check that can be bypassed, or a weak derivation step, any of which would let the server accept a crafted request as a legitimate user's. This is our inference from the classification, not something Fortinet has confirmed.

CVSS vector

Fortinet, acting as CNA, supplied the score of 7.5 (High). NVD's own assessment is shown as unavailable, so no one has independently corroborated the score.

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
  • AV:N: network attack vector
  • AC:L: low attack complexity
  • PR:N: no privileges required
  • UI:N: no user interaction
  • S:U: scope unchanged
  • C:H: high confidentiality impact
  • I:N / A:N: no integrity or availability impact in the published assessment

Attack flow, as far as the disclosure supports it

  1. The attacker needs network reachability to the affected FortiMail API (AV:N) and no credentials (PR:N).
  2. The attacker sends a crafted request. No victim interaction is needed (UI:N), and Fortinet rates the complexity as low (AC:L).
  3. Because the authentication algorithm is implemented incorrectly, the appliance treats the crafted request as authenticated for an account the attacker does not legitimately control (inferred from CWE-303 and the NVD description).
  4. The result described by NVD is account takeover, which the vector scores as high confidentiality impact.

We cannot go further than this without speculating. The available material does not establish administrator takeover, code execution, or a specific protocol failure.

Advisory inconsistencies

The FG-IR-26-203 page has two discrepancies you should know about before you escalate internally:

  • Product naming: the title references "fortiweb2" and the summary says FortiWeb, while the affected release and solution table lists FortiMail. NVD and the CVE description also name FortiMail. This reads as a templating mistake. Treat it as a FortiMail advisory, and do not assume FortiWeb is affected by this CVE.
  • Impact field: the advisory's impact field says "Denial of service", while its summary and NVD describe account takeover. The vector assigns no availability impact. Do not combine the two into a claim that both outcomes have been demonstrated.

Exploitation status

Fortinet marks this issue "Known Exploited: No". VulDB, a secondary database, also reports "KEV: no". No threat actor is named, no working exploit or request sample appears in the primary records, and the advisory provides no indicators of compromise. "No known exploitation" is not proof that exploitation cannot occur.

Timeline

  • Early October 2026: FortiMail 8.0.2, 7.6.7 and 7.4.9 released as part of a coordinated security hardening release
  • October 5, 2026: FortiMail Cloud remediated
  • November 10, 2026: Fortinet advisory FG-IR-26-203 published
  • November 11, 2026: NVD publication

Patch Information

Where the fix lives. The fix ships only as firmware builds from Fortinet. Per FG-IR-26-203:

  • FortiMail 8.0: 8.0.0 through 8.0.1 vulnerable, fixed in 8.0.2 or above
  • FortiMail 7.6: 7.6.0 through 7.6.6 vulnerable, fixed in 7.6.7 or above
  • FortiMail 7.4: 7.4.0 through 7.4.8 vulnerable, fixed in 7.4.9 or above
  • FortiMail 7.2: all versions (7.2.0 through 7.2.9) vulnerable, with no fix in this branch. Migrate to a fixed release.

There is no backport for 7.2. Fortinet's October hardening release post gives the instruction for 7.2.0 through 7.2.9 as "Upgrade to 7.4 branch or later". For 7.2 customers, remediation means a major version migration rather than a maintenance update.

Fixed first, disclosed later. Fortinet shipped the fixed builds in early October 2026 and deliberately held the advisories until its November 10 Patch Tuesday. Fortinet says the delay gives customers time to upgrade before technical details are public, and notes that AI-assisted patch diffing can now turn a fix into a working exploit in hours or days.

Build numbers. A third-party write-up citing Fortinet's documentation change log gives the fixed builds as:

  • 7.6.7 build 858 (release notes dated October 2, 2026)
  • 8.0.2 build 263 (October 3, 2026)
  • 7.4.9 build 630

Checking the build number is a quick way to confirm whether an appliance carries the fix.

The release notes will not tell you. The 8.0.2 and 7.6.7 "Resolved issues" pages list fixed classes such as CWE-22, CWE-89, CWE-321, CWE-522, CWE-472 and CWE-1050. Neither lists CWE-303 or any authentication algorithm item. Teams that assess exposure from release-note CWE tables would miss this issue. The advisory is the only public document tying CVE-2026-126151 to these builds.

Other notes:

  • The advisory lists Virtual Patch: No, so no FortiGuard IPS or WAF signature covers this flaw, and there is no workaround. Firmware is the only vendor remedy.
  • FortiMail Cloud is already fixed. Fortinet updated all FortiMail Cloud instances to 7.6.7 GA or 8.0.2 GA on October 5, 2026, and states that FortiMail Cloud customers need not take action. That exception applies only to FortiMail Cloud.
  • If the fix corrects the server-side authentication or verification logic, as the CWE label suggests, it cannot be delivered as a configuration change, which would explain why it ships as firmware. Again, this is inference.
  • HKCERT includes this CVE in its Fortinet bulletin and directs readers to the vendor fixes. It does not add a separate workaround or version matrix.

Affected Systems and Versions

  • FortiMail 8.0.0 through 8.0.1
  • FortiMail 7.6.0 through 7.6.6
  • FortiMail 7.4.0 through 7.4.8
  • FortiMail 7.2, all versions. NVD's affected products table represents this as 7.2.0 through 7.2.9, but both the NVD description and Fortinet's advisory say all 7.2 versions. The narrower database range is not evidence of a safe 7.2 release.
  • FortiMail Cloud, remediated by Fortinet on October 5, 2026. No customer action is required.

The available sources do not identify a specific configuration that is or is not exposed beyond the version ranges above.

Vendor Security History

This issue was found internally. Fortinet credits Gwendal Guegniaud of its Product Security team, using Claude Mythos, as part of the AI-assisted testing program behind the October 2026 hardening releases. Fortinet's vulnerability policy states that High severity issues are addressed in supported versions through its monthly advisory process, with accelerated publication possible when observed exploitation increases customer risk.

Fortinet products also have a documented exploitation history:

  • FortiWeb CVE-2025-64446 and CVE-2025-58034: CISA documented exploitation, describing potential initial access through the first and privilege escalation by chaining the second, and both were added to the Known Exploited Vulnerabilities catalog in November 2025. This is unrelated to the FortiWeb wording on this CVE's advisory.
  • FortiMail CVE-2026-104286: on October 6, 2026, Singapore's CSA reported active exploitation of this separate FortiMail flaw, a path traversal permitting unauthenticated arbitrary file writing through crafted HTTP or HTTPS requests, scored CVSS 3.1 9.8.

The mechanisms and evidence differ. CVE-2026-104286 is an actively exploited path traversal; CVE-2026-126151 is an authentication implementation flaw with no known exploitation. Do not apply indicators and attack details from the older alert to this CVE.

References

Detect & fix
what others miss

Works with
  • GitHub
  • GitLab
  • Bitbucket
  • Azure DevOps Services
  • Gerrit
  • CVS
  • Jira
  • Linear
  • Slack
  • Security Compass
Explore integrations
Security magnifying glass visualization