Security readout for executives and security teams
Plain-English summary
MLflow deployments using artifact serving can let an authenticated user write over another user’s uploaded artifacts. In ML systems, that can corrupt trusted model files and create downstream execution risk when poisoned models are later loaded. The source bundle lists a critical CVSS 9.0 score and says the issue is fixed in MLflow 3.10.0.
Executive priority
Treat this as urgent for ML platforms that share MLflow across users or feed artifacts into production. Prioritize upgrade and exposure review before relying on affected artifact repositories for model deployment decisions.
Technical view
The flaw is missing resource-level authorization on /mlflow-artifacts/mpu/* multipart upload endpoints when --serve-artifacts is enabled. A low-privileged authenticated attacker may perform cross-user artifact writes. Reported impacts include model supply-chain poisoning and arbitrary code execution when compromised artifacts are consumed. A commit reference and vendor entries are provided, but detailed deployment constraints are limited.
Likely exposure
Exposure appears limited to MLflow deployments on affected versions with --serve-artifacts enabled and reachable MPU artifact endpoints. Multi-user environments are higher risk because the core issue is cross-user write authorization. The bundle does not prove all MLflow installations are affected.
Exploitation context
The bundle does not cite active exploitation, and KEV is false. Exploitation requires authenticated access according to the CVSS vector, with user interaction marked required. The practical risk is highest where model artifacts are trusted and automatically loaded into production or research workflows.
Researcher notes
The available evidence identifies CWE-862 and CWE-1220, affected MPU routes, --serve-artifacts dependency, and a fixing commit. The source bundle has a version inconsistency: affected versions include <=3.10.1.dev0 while resolution is 3.10.0. Confirm exact version semantics from upstream before broad fleet conclusions.
Mitigation direction
- Upgrade MLflow to version 3.10.0 or later per the source bundle.
- Check MLflow vendor guidance for any version-specific deployment instructions.
- Restrict access to MLflow artifact endpoints to trusted authenticated users.
- Disable or avoid --serve-artifacts until upgraded, if operationally feasible.
- Review model artifact provenance before loading models from affected systems.
Validation and detection
- Inventory MLflow versions across development, staging, and production systems.
- Confirm whether MLflow is started with --serve-artifacts enabled.
- Review logs for access to /mlflow-artifacts/mpu/ endpoints.
- Check whether untrusted users had MLflow access during the exposure window.
- Verify upgraded systems are running the fixed MLflow version.
Public sources used
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
CWE-1220: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCWE-862: Authorization and privilege behavior lookup
Authorization weaknesses can support privilege escalation and valid-account review, depending on exploit path. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupExecution behavior lookup
The CVE wording references code or command execution, so execution technique review may help defensive triage. This is a Glexia inferred lookup path, not an official MITRE, ATT&CK, or CVE Program mapping.
Open ATT&CK lookupCVE-2026-2651 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
Open ATT&CK lookup- Severity
- Critical
- CVSS
- 9 (3.0)
- Known Exploited
- No
- Published
Vector: CVSS:3.0/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H
CNA and ADP enrichment extracted from CVE v5
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
SSVC decision data
CVSS vector scores
2 official scoresWe collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CVSS:3.0/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H2.36@huntr_aiCVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H2.36redhat-SADPVulnerability scoring details
Base CVSS 3.1 score
9CriticalVector: CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H
Base CVSS 3.0 score
9CriticalVector: CVSS:3.0/AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H
Vulnerability timeline
Timeline events are normalized from CVE metadata, CNA source timelines, ADP timelines, and KEV metadata when present.
- CVE reservedCVE Program
The CVE ID was reserved by the assigning CNA.
- ADP timelineredhat-SADP
Made public.
- CVE publishedCVE Program
The CVE record was published.
- ADP timelineredhat-SADP
Reported to Red Hat.
- CVE updatedCVE Program
The CVE record metadata indicates this as the latest update time.
ADP provider summaries
- 2026-05-25T07:00:55.073Z: Reported to Red Hat.
- 2026-05-25T06:00:34.002Z: Made public.
Source materials
- CVE List V5 sourceCVE List V5
- https://huntr.com/bounties/65beb119-d3e0-4e03-af2f-fa98f78f83dcCVE reference, @huntr_ai
- https://github.com/mlflow/mlflow/commit/d7290811d8f3c95366d80109424edc1fb1ad966fCVE reference, @huntr_ai
- https://access.redhat.com/security/cve/CVE-2026-2651CVE reference, redhat-SADP · vdb-entry, x_refsource_REDHAT
- RHBZ#2481117CVE reference, redhat-SADP · issue-tracking, x_refsource_REDHAT
- https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-2651.jsonCVE reference, redhat-SADP · x_sadp-csaf-vex
Products and packages named in the record
CWE details
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.
Insufficient Granularity of Access Control
Insufficient Granularity of Access Control represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.
Missing Authorization
Missing Authorization represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.