Cyber Resilience

CVE-2025-46722

Vllm 0.7.0 – 0.9.0

Published
29 May 2025
Modified
17 June 2026
Patch / advisory
View advisory
CVSS Score v3.1 4.2
Click a component to see what it means
Raw vectorCVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:L
EPSS Score 0.0028 20th percentile
Risk Priority 34 floored blend · peak EPSS

Summary

CVE-2025-46722 is a medium-severity Incomplete Comparison with Missing Factors (CWE-1023) vulnerability in Vllm Vllm. Its CVSS base score is 4.2 (Medium).

Operationally, exploitation aligns with the MITRE ATT&CK technique Masquerading (T1036); ranked at the 20th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog.

This vulnerability is AI-related — categorised as NLP and Transformers; in the Data-Related Vulnerabilities risk domain.

The strongest mitigations our analysis identified map to SI-10 (Information Input Validation) and AC-24 (Access Control Decisions) — see the control section below for these in your framework.

EU & UK References

Vulnerability Data

vLLM is an inference and serving engine for large language models (LLMs). In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently,…

more

it serializes PIL.Image.Image objects using only obj.tobytes(), which returns only the raw pixel data, without including metadata such as the image’s shape (width, height, mode). As a result, two images of different sizes (e.g., 30x100 and 100x30) with the same pixel byte sequence could generate the same hash value. This may lead to hash collisions, incorrect cache hits, and even data leakage or security risks. This issue has been patched in version 0.9.0.

CWE(s)
CWE-1023CWE-1288

AI Security AnalysisAI

AI Category
NLP and Transformers
Risk Domain
Data-Related Vulnerabilities
OWASP Top 10 for LLMs 2025
None mapped
Classification Reason
Matched keywords: llms, vllm

Related Threats

MITRE ATT&CK Enterprise Techniques

T1036 Masquerading Stealth
Adversaries may attempt to manipulate features of their artifacts to make them appear legitimate or benign to users and/or security tools.
attack.mitre.org →
T1190 Exploit Public-Facing Application Initial Access
Adversaries may attempt to exploit a weakness in an Internet-facing host or system to initially access a network.
attack.mitre.org →
T1203 Exploitation for Client Execution Execution
Adversaries may exploit software vulnerabilities in client applications to execute code.
attack.mitre.org →
T1212 Exploitation for Credential Access Credential Access
Adversaries may exploit software vulnerabilities in an attempt to collect credentials.
attack.mitre.org →
T1553 Subvert Trust Controls Defense Impairment
Adversaries may undermine security controls that will either warn users of untrusted activity or prevent execution of untrusted programs.
attack.mitre.org →
Derived from this CVE’s CWE(s) via the direct CWE→ATT&CK cross-walk.

CVEs Like This One

CVE-2026-44223Same product: Vllm Vllm
CVE-2025-62372Same product: Vllm Vllm
CVE-2026-44222Same product: Vllm Vllm
CVE-2024-11041Same product: Vllm Vllm
CVE-2025-24357Same product: Vllm Vllm
CVE-2025-47277Same product: Vllm Vllm
CVE-2025-30165Same product: Vllm Vllm
CVE-2025-32444Same product: Vllm Vllm
CVE-2025-29783Same product: Vllm Vllm
CVE-2025-62164Same product: Vllm Vllm

Affected Assets

vllm
vllm
0.7.0 — 0.9.0

Mitigating Controls

Control response

Prevent
Stop it (NIST 800-53)

—

Detect
Catch it (NIST detect / respond)

—

Harden
Shrink the surface (DISA STIG)

—

Validate
Prove the fix (OWASP ASVS)
  • V6.3.4
  • V6.3.5
  • V6.1.3
  • V6.5.7

Mitigating Controls (NIST 800-53 r5) AI

SI-10 Information Input Validation good match

SI-10 requires validity checks on supplied inputs, directly stopping acceptance of internally inconsistent complex data.

AC-24 Access Control Decisions partial match

Mandating that access-control decisions apply the full set of required rules to each request structurally eliminates missing-factor comparisons.

AC-3 Access Enforcement partial match

Access enforcement requires that every authorization decision evaluate all relevant entity attributes, directly stopping incomplete comparisons from being used.

Mitigating Controls (NIST CSF 2.0) AI

Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→CSF cross-walk (authority under review) — links open the control.

PR.PS-06 mostly match
prevents

Secure SDLC practices (reviews, testing, static analysis) directly prevent incomplete comparison flaws from being introduced.

PR.AA-03 partial match
prevents

Authentication decisions often rely on multi-factor comparisons; incomplete comparisons directly weaken this control.

PR.AA-05 partial match
prevents

Enforcing access policy requires complete evaluation of all relevant attributes; missing factors undermine authorization decisions.

Mitigating Controls (ISO/IEC 27001:2022 Annex A) AI

Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→ISO cross-walk (authority under review) — links open the control.

A.8.29 Security testing in development and acceptance mostly match
finds

Security testing can detect missing comparison factors, yet testing is only one part of the control's scope.

A.8.25 Secure development life cycle partial match
prevents

Secure development lifecycle requires input validation and consistency checks that directly address CWE-1288.

A.8.26 Application security requirements partial match
prevents

Application security requirements can mandate complete multi-factor comparisons, but the control is broader than this single weakness.

A.8.27 Secure system architecture and engineering principles partial match
prevents

Secure architecture principles can require exhaustive entity comparisons, yet the control addresses many other design concerns.

A.8.28 Secure coding partial match
prevents

Secure coding standards can forbid incomplete comparisons, but the control covers a wide range of coding issues.

References