LiveActive security incident?Get immediate response
CVE Record

CVE-2026-22778: vLLM leaks a heap address when PIL throws an error

vLLM is an inference and serving engine for large language models (LLMs). From 0.8.3 to before 0.14.1, when an invalid image is sent to vLLM's multimodal endpoint, PIL throws an error. vLLM returns this error to the client, leaking a heap address. With this leak, we reduce ASLR from 4 billion guesses to ~8 guesses. This vulnerability can be chained a heap overflow with JPEG2000 decoder in OpenCV/FFmpeg to achieve remote code execution. This vulnerability is fixed in 0.14.1.

CriticalCVSS 9.8Not KEV-listedUpdated
Discuss exposureBack to CVEs
Glexia's TakeAutomated analysiscritical

Security readout for executives and security teams

Plain-English summary

Affected vLLM servers can reveal internal memory addresses when they receive an invalid image. That leak weakens a key protection, ASLR, and the advisory says it can help chain another image-decoder flaw into remote code execution. The fixed version is vLLM 0.14.1.

Executive priority

Treat this as urgent for AI services that process images. The direct flaw is an information leak, but the vendor describes it as a critical enabler for bypassing memory protections and chaining to code execution.

Technical view

vLLM versions 0.8.3 through before 0.14.1 return PIL image-processing errors to clients from the multimodal endpoint, exposing a heap address. The advisory rates this CVSS 9.8 and links the disclosure to reduced ASLR entropy and possible chaining with a JPEG2000 decoder heap overflow.

Likely exposure

Highest exposure is vLLM deployments on versions >=0.8.3 and <0.14.1 that accept image input through multimodal endpoints, especially externally reachable APIs. Text-only deployments or systems without untrusted image input appear less exposed, but version and route configuration must be verified.

Exploitation context

The source bundle does not cite KEV listing or confirmed active exploitation. It does state unauthenticated network attack conditions and a plausible chain to remote code execution when combined with a separate image-decoder heap overflow.

Researcher notes

Focus triage on affected vLLM versions and multimodal request paths. Evidence supports the heap-address leak and fixed release, but the bundle does not prove active exploitation or define a standalone RCE without an additional decoder vulnerability.

Mitigation direction

  • Upgrade vLLM to 0.14.1 or later.
  • Apply relevant Red Hat advisories where Red Hat-packaged products are in use.
  • Restrict untrusted access to vLLM multimodal image endpoints until upgraded.
  • Check vendor guidance for any environment-specific mitigations or backports.
  • Avoid exposing diagnostic exception details to external clients.

Validation and detection

  • Inventory all vLLM deployments and package versions.
  • Confirm no deployed version is >=0.8.3 and <0.14.1.
  • Identify whether multimodal image endpoints accept untrusted client input.
  • Review API exposure for internet-facing or unauthenticated vLLM services.
  • Check vendor advisories for fixed Red Hat package status.
Prepared
Confidence
high
Sources
11

Public sources used

Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.

Potential ATT&CK relevance

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 · medium confidence lookup

CWE-209: Information exposure and cloud metadata lookup

Information exposure and SSRF weaknesses can make discovery, cloud metadata, and credential material review relevant. 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 lookup
cwe · low confidence lookup

CWE-532: 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 lookup
description · low confidence lookup

Execution 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 lookup
cve · low confidence lookup

CVE-2026-22778 mapping review

Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.

Open ATT&CK lookup
Vulnerability profileCVE Program record
Severity
Critical
CVSS
9.8 (3.1)
Known Exploited
No
Published

Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

Official CVE source material

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.

2CVSS vectors
5Timeline events
2ADP providers
16Source links

SSVC decision data

CISA-ADPCISA Coordinator
Timestamp
Version
2.0.3
Exploitation: noneAutomatable: yesTechnical Impact: total

CVSS vector scores

2 official scores

We 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.

ScoreVersionSeverityVectorExploitImpactSource
9.8CVSS 3.1CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H3.95.9GitHub_M
9.8CVSS 3.1CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H3.95.9redhat-SADP

Vulnerability scoring details

Base CVSS 3.1 score

9.8Critical
CVSS 3.1 vector shape for CVE-2026-22778Attack VectorAttack ComplexityPrivileges RequiredUser InteractionScopeConfidentiality ImpactIntegrity ImpactAvailability Impact

Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

Attack Vector
NetworkAdjacentLocalPhysical
Attack Complexity
LowHigh
Privileges Required
NoneLowHigh
User Interaction
NoneRequired
Scope
ChangedUnchanged
Confidentiality Impact
HighLowNone
Integrity Impact
HighLowNone
Availability Impact
HighLowNone

Vulnerability timeline

Timeline events are normalized from CVE metadata, CNA source timelines, ADP timelines, and KEV metadata when present.

  1. CVE reservedCVE Program

    The CVE ID was reserved by the assigning CNA.

  2. ADP timelineredhat-SADP

    Made public.

  3. CVE publishedCVE Program

    The CVE record was published.

  4. ADP timelineredhat-SADP

    Reported to Red Hat.

  5. CVE updatedCVE Program

    The CVE record metadata indicates this as the latest update time.

ADP provider summaries

CISA-ADPCISA ADP Vulnrichment
other:ssvc
redhat-SADPvLLM: vLLM: Remote code execution via invalid image processing in the multimodal endpoint.
other:Red Hat severity ratingcvssV3_1
  • 2026-02-03T00:01:43.512Z: Reported to Red Hat.
  • 2026-02-02T21:09:53.265Z: Made public.

Source materials

Affected products

Products and packages named in the record

VendorProductVersion / packageStatus
vllm-projectvllm>= 0.8.3, < 0.14.1Listed
Weakness

CWE details

CWE-209CWE-532

CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.

CWE-209 · source CWE mapping

Generation of Error Message Containing Sensitive Information

Generation of Error Message Containing Sensitive Information represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.

Open CWE intelligenceSame-CWE CVEs
CWE-532 · source CWE mapping

Insertion of Sensitive Information into Log File

Insertion of Sensitive Information into Log File represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.

Open CWE intelligenceSame-CWE CVEs