CVE-2026-73558: vLLM Cross-User Data Leak Vulnerability Vulnerability

MODERATE
Advisory Identifier: GHSA-7m6h-x95x-82q5
Share
CVSS 5.3
CWE-190
Published
Advisory Summary

CVE-2026-73558 is a MODERATE (CVSS 5.3) cross-user data leak vulnerability vulnerability (CWE-190: Integer Overflow or Wraparound) affecting vLLM libraries in the PyPI ecosystem. The issue involves cross-user data leak vulnerability. Remediation is available by upgrading affected packages to version 0.27.0.

CVE-2026-73558 at a Glance

CVE Identifier CVE-2026-73558
GHSA Identifier GHSA-7m6h-x95x-82q5
Severity MODERATE
CVSS v3.1 Base Score 5.3
CWE Classification CWE-190: Integer Overflow or Wraparound
Affected Product vLLM
Package Ecosystem PyPI
Published Date
Last Updated

Security Impact

CVSS: 5.3

The security impact metrics below are parsed directly from the CVSS v3.1 vector string (CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N).

Attack Vector Network (AV:N)
Attack Complexity High (AC:H)
Privileges Required None (PR:N)
User Interaction Required (UI:R)
Scope Unchanged (S:U)
Confidentiality High (C:H)
Integrity None (I:N)
Availability None (A:N)

How to Fix CVE-2026-73558

To remediate this vulnerability, upgrade all affected vLLM dependencies to a patched release. Both direct and transitive dependencies should be checked, as frameworks or security libraries may pull in vulnerable versions transitively.

vllm fixed: 0.27.0 
pip install vllm>=0.27.0
Verify Direct and Transitive Dependencies

Run the following command in your project directory to verify that all resolved dependencies use the patched release: 

pip show vllm

How to Check if You Are Affected

To determine whether your Python environment is affected by CVE-2026-73558, inspect your installed packages using pip:

pip show vllm

Analyze the terminal output to locate any version of vLLM in your dependency tree. If any resolved version matches the vulnerable ranges listed in the table below (for example, versions earlier than 0.27.0), your application is affected.

Note: The pip show vllm command checks currently installed packages in your active Python environment.

Technical Details

Vulnerability Mechanism & Technical Analysis

This vulnerability is classified under CWE-190: Integer Overflow or Wraparound, where the product performs a calculation that can produce an integer overflow or wraparound when the logic assumes that the resulting value will always be larger than the original value. this occurs when an integer value is incremented to a value that is too large to store in the associated representation. when this occurs, the value may become a very small or negative number. In vLLM, software state validation or input boundaries are circumvented during processing. Review the official source advisory statement below for specific affected components, classes, and reproduction details.

Source Advisory Statement

### Summary An integer overflow in the act_and_mul_kernel kernel can cause the output of one user request to be incorporated into the response of another request within the same inference batch. Under certain conditions, the last request in a batch can receive a partial or complete copy of the first user's inference result, resulting in cross-user data leakage. ### Details The root cause is an integer overflow in the expression blockIdx.x * 2 * d at https://github.com/vllm-project/vllm/blob/ff712f6447093d07747c88680b9d006b119f5890/csrc/activation_kernels.cu#L82. As a result, the computation for one user (User A) can incorrectly consume input data from another user (User B). In particular, when 2^32 is divisible by d, the overflow can cause User A's output to contain portions of User B's inference result. In some cases, User B's response may be copied entirely into User A's response. This constitutes a severe cross-user information disclosure vulnerability and is straightforward to trigger. ### PoC We reproduced the issue using meta-llama/Llama-3.2-1B-Instruct, for which d = 8192. Using the following configuration: Batch size: 17 Sequence length: 16384 The final response in the batch becomes an exact copy of the first response in the batch, demonstrating complete cross-user data leakage. ### Impact This vulnerability enables cross-user information disclosure. An attacker can intentionally craft requests that are processed within the same inference batch as a victim's request and cause the victim's inference output to be copied into the attacker's response. As a result, sensitive information contained in another user's model response may be exposed to an unauthorized party. ### Versions For versions prior and equal to 0.21.0, the bug is in csrc/activation_kernels.cu, and for versions later than 0.21.0, the bug is in csrc/libtorch_stable/activation_kernels.cu.

CVE-2026-73558 Affected vLLM Versions

1 component(s)
Fixed Versions:
0.27.0
Product / Package Ecosystem Affected Versions Fixed Version
vllm
pypi/vllm
PyPI
< 0.27.0
0.27.0

References

GitHub Advisory / Wiki ADVISORY
https://github.com/vllm-project/vllm/security/advisories/GHSA-7m6h-x95x-82q5
NVD — CVE-2026-73558 NVD
https://nvd.nist.gov/vuln/detail/CVE-2026-73558
vLLM Source Repository REPOSITORY
https://github.com/vllm-project/vllm/issues/42860
vLLM Source Repository REPOSITORY
https://github.com/vllm-project/vllm/pull/49660
Security Fix Commit FIX COMMIT
https://github.com/vllm-project/vllm/commit/451227cb3ff07989698fed982c2d3e4300257924
vLLM Source Repository REPOSITORY
https://github.com/vllm-project/vllm
vLLM Release RELEASE
https://github.com/vllm-project/vllm/releases/tag/v0.27.0

Frequently Asked Questions

What is CVE-2026-73558?
CVE-2026-73558 is a security vulnerability in vLLM involving Cross-User Data Leak Vulnerability under CWE-190: Integer Overflow or Wraparound.
Which vLLM versions are affected?
Affected versions include releases prior to 0.27.0 across affected packages: vllm.
What version fixes CVE-2026-73558?
Upgrading to 0.27.0 (or the latest patched release for your respective branch) resolves the vulnerability.
Is vllm affected?
Yes, vllm is affected in releases prior to 0.27.0. Upgrade your dependency to 0.27.0 or later.
What is the CVSS score of CVE-2026-73558?
CVE-2026-73558 has a CVSS v3.1 base score of 5.3 (MODERATE) with vector string CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N.
What CWE is associated with CVE-2026-73558?
CVE-2026-73558 is classified under CWE-190: Integer Overflow or Wraparound (The product performs a calculation that can produce an integer overflow or wraparound when the logic assumes that the resulting value will always be larger than the original value. This occurs when an integer value is incremented to a value that is too large to store in the associated representation. When this occurs, the value may become a very small or negative number.).
How can I check PyPI dependencies for CVE-2026-73558?
Run the dependency tree command: 'pip show vllm'. This command lists direct and transitive dependencies in your build configuration so you can verify if a vulnerable version is resolved.
Machine-readable OSV JSON OSV v1.5 Schema
Normalized schema output
View Raw JSON 
{
  "schema_version" : "1.4.0",
  "id" : "GHSA-7m6h-x95x-82q5",
  "modified" : "2026-09-08T20:24:49Z",
  "published" : "2026-09-08T20:24:49Z",
  "aliases" : [ "CVE-2026-73558" ],
  "summary" : "vLLM: Cross-User Data Leak Vulnerability",
  "details" : "### Summary\nAn integer overflow in the act_and_mul_kernel kernel can cause the output of one user request to be incorporated into the response of another request within the same inference batch. Under certain conditions, the last request in a batch can receive a partial or complete copy of the first user's inference result, resulting in cross-user data leakage.\n\n\n### Details\nThe root cause is an integer overflow in the expression blockIdx.x * 2 * d at https://github.com/vllm-project/vllm/blob/ff712f6447093d07747c88680b9d006b119f5890/csrc/activation_kernels.cu#L82. \n\nAs a result, the computation for one user (User A) can incorrectly consume input data from another user (User B). In particular, when 2^32 is divisible by d, the overflow can cause User A's output to contain portions of User B's inference result. In some cases, User B's response may be copied entirely into User A's response.\n\nThis constitutes a severe cross-user information disclosure vulnerability and is straightforward to trigger.\n### PoC\nWe reproduced the issue using meta-llama/Llama-3.2-1B-Instruct, for which d = 8192.\n\nUsing the following configuration:\n\nBatch size: 17\nSequence length: 16384\n\nThe final response in the batch becomes an exact copy of the first response in the batch, demonstrating complete cross-user data leakage.\n\n### Impact\nThis vulnerability enables cross-user information disclosure. An attacker can intentionally craft requests that are processed within the same inference batch as a victim's request and cause the victim's inference output to be copied into the attacker's response.\n\nAs a result, sensitive information contained in another user's model response may be exposed to an unauthorized party.\n\n### Versions\n\nFor versions prior and equal to 0.21.0, the bug is in csrc/activation_kernels.cu, and for versions later than 0.21.0, the bug is in csrc/libtorch_stable/activation_kernels.cu.",
  "severity" : [ {
    "type" : "CVSS_V3",
    "score" : "CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:N/A:N"
  } ],
  "affected" : [ {
    "package" : {
      "ecosystem" : "PyPI",
      "name" : "vllm"
    },
    "ranges" : [ {
      "type" : "ECOSYSTEM",
      "events" : [ {
        "introduced" : "0"
      }, {
        "fixed" : "0.27.0"
      } ]
    } ]
  } ],
  "references" : [ {
    "type" : "WEB",
    "url" : "https://github.com/vllm-project/vllm/security/advisories/GHSA-7m6h-x95x-82q5"
  }, {
    "type" : "ADVISORY",
    "url" : "https://nvd.nist.gov/vuln/detail/CVE-2026-73558"
  }, {
    "type" : "WEB",
    "url" : "https://github.com/vllm-project/vllm/issues/42860"
  }, {
    "type" : "WEB",
    "url" : "https://github.com/vllm-project/vllm/pull/49660"
  }, {
    "type" : "WEB",
    "url" : "https://github.com/vllm-project/vllm/commit/451227cb3ff07989698fed982c2d3e4300257924"
  }, {
    "type" : "PACKAGE",
    "url" : "https://github.com/vllm-project/vllm"
  }, {
    "type" : "WEB",
    "url" : "https://github.com/vllm-project/vllm/releases/tag/v0.27.0"
  } ],
  "database_specific" : {
    "cwe_ids" : [ "CWE-190" ],
    "severity" : "MODERATE",
    "github_reviewed" : true,
    "github_reviewed_at" : "2026-09-08T20:24:49Z",
    "nvd_published_at" : "2026-08-13T15:20:18Z"
  }
}