CVE-2026-53923
CVE-2026-53923 is a high-severity vulnerability in Vllm with a CVSS 3.x base score of 7.5. It is not currently listed as actively exploited by CISA, and its EPSS exploit-prediction score is low. The underlying weakness is classified as CWE-200.
Key facts
- Severity: High (CVSS 3.x base score 7.5)
- CVSS v4: 5.3
- EPSS exploit prediction: 0% (20th percentile)
- Actively exploited: Not listed in CISA KEV
- EU (EUVD) id: EUVD-2026-38400
- Weakness: CWE-200
- Affected product: Vllm
- Published:
- Last modified:
Description
vLLM is an inference and serving engine for large language models (LLMs). From 0.5.5 until 0.23.1rc0, integer truncation of tensor dimensions in vLLM's GGUF dequantize kernels (csrc/quantization/gguf/gguf_kernel.cu) causes partial tensor processing. The output tensor is allocated at full size via torch::empty (uninitialized memory), but the dequantize CUDA kernel processes only a truncated number of elements. The unfilled portion of the output tensor retains whatever was previously in GPU memory. In multi-tenant inference deployments, this residual GPU memory may contain tensor data from other users' inference requests, constituting information disclosure. This vulnerability is fixed in 0.23.1rc0.
Frequently asked questions
- What is CVE-2026-53923?
- vLLM is an inference and serving engine for large language models (LLMs). From 0.5.5 until 0.23.1rc0, integer truncation of tensor dimensions in vLLM's GGUF dequantize kernels (csrc/quantization/gguf/gguf_kernel.cu) causes partial tensor processing. The output tensor is allocated at full size via torch::empty (uninitialized memory), but the dequantize CUDA kernel processes only a truncated number of elements. The unfilled portion of the output tensor retains whatever was previously in GPU memory. In multi-tenant inference deployments, this residual GPU memory may contain tensor data from other users' inference requests, constituting information disclosure. This vulnerability is fixed in 0.23.1rc0.
- How severe is CVE-2026-53923?
- CVE-2026-53923 has a CVSS 3.x base score of 7.5, rated high severity. It is exploitable over network with low attack complexity, requires no privileges and no user interaction. Impact on confidentiality is high, integrity none, and availability none.
- Is CVE-2026-53923 being actively exploited?
- It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 0% (20th percentile), an estimate of the probability of exploitation in the next 30 days.
- What products are affected by CVE-2026-53923?
- CVE-2026-53923 affects Vllm. See the affected-products list for the exact vulnerable versions.
- How do I fix CVE-2026-53923?
- Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround. Given its high severity, prioritise patching exposed systems.
- Does CVE-2026-53923 have an EU (EUVD) identifier?
- Yes. CVE-2026-53923 is tracked in the ENISA EU Vulnerability Database (EUVD) as EUVD-2026-38400.
- When was CVE-2026-53923 published?
- CVE-2026-53923 was published on 2026-06-22 and last updated on 2026-06-24.
References
- https://github.com/vllm-project/vllm/commit/f219788f91952827132fa4fdf916427cd20d225e
- https://github.com/vllm-project/vllm/pull/44971
- https://github.com/vllm-project/vllm/security/advisories/GHSA-5jv2-g5wq-cmr4
Affected products (1)
- cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*
More vulnerabilities in Vllm
- CVE-2025-32444 — Critical (CVSS 10.0): vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Versions starting from 0.6.5 and…
- CVE-2026-22778 — Critical (CVSS 9.8): vLLM is an inference and serving engine for large language models (LLMs). From 0.8.3 to before 0.14.1, when an invalid…
- CVE-2025-47277 — Critical (CVSS 9.8): vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4…
- CVE-2024-11041 — Critical (CVSS 9.8): vllm-project vllm version v0.6.2 contains a vulnerability in the MessageQueue.dequeue() API function. The function uses…
- CVE-2026-48746 — Critical (CVSS 9.1): vLLM is an inference and serving engine for large language models (LLMs). From 0.3.0 until 0.22.0, a vulnerability in…
- CVE-2025-29783 — Critical (CVSS 9.0): vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. When vLLM is configured to use…
Other CWE-200 (Exposure of Sensitive Information to an Unauthorized Actor) vulnerabilities
- CVE-2026-27604 — Critical (CVSS 10.0): FOSSBilling is a free, open-source billing and client management system. Starting in version 0.5.4 and prior to version…
- CVE-2026-40965 — Critical (CVSS 10.0): Cloud Foundry UAA versions v76.12.0 through v78.12.0 are vulnerable to a private key exposure. The server contains a…
- CVE-2026-42826 — Critical (CVSS 10.0): Exposure of sensitive information to an unauthorized actor in Azure DevOps allows an unauthorized attacker to disclose…
- CVE-2025-29270 — Critical (CVSS 10.0): Incorrect access control in the realtime.cgi endpoint of Deep Sea Electronics devices DSE855 v1.1.0 to v1.1.26 allows…
- CVE-2025-61481 — Critical (CVSS 10.0): An issue in MikroTik RouterOS v.7.14.2 and SwOS v.2.18 exposes the WebFig management interface over cleartext HTTP by…
- CVE-2025-53624 — Critical (CVSS 10.0): The Docusaurus gists plugin adds a page to your Docusaurus instance, displaying all public gists of a GitHub user.…
Browse all CWE-200 (Exposure of Sensitive Information to an Unauthorized Actor) vulnerabilities →