# CVE-2026-22778

## Summary

- **CVE ID:** CVE-2026-22778
- **Severity:** CRITICAL
- **CVSS Score:** 9.8 (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H)
- **CWE:** CWE-532
- **Published:** Feb 2, 2026
- **Last Modified:** Jul 15, 2026

## Description

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.

## Affected Products

- vllm-project — vllm (>= 0.8.3, < 0.14.1)

## References

- [CNA](https://github.com/vllm-project/vllm/security/advisories/GHSA-4r2x-xpjr-7cvv)
- [CNA](https://github.com/vllm-project/vllm/pull/31987)
- [CNA](https://github.com/vllm-project/vllm/pull/32319)
- [CNA](https://github.com/vllm-project/vllm/releases/tag/v0.14.1)
- [redhat-SADP](https://access.redhat.com/security/cve/CVE-2026-22778)
- [redhat-SADP](https://bugzilla.redhat.com/show_bug.cgi?id=2436113)
- [redhat-SADP](https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-22778.json)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:3461)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:3462)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:3782)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:19712)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:3713)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:30089)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:30088)
- [redhat-SADP](https://access.redhat.com/errata/RHSA-2026:30087)

## Exploitation Prediction (EPSS)

- **EPSS Score:** 3.72%
- **EPSS Percentile:** 89.1

---
_Exported from OnDuty AI Vulnerability Intelligence on 2026-09-11._