Skip to content
securitytabs

CVE-2026-28400

High7.5

model-runner: Docker Model Runner Unauthenticated Runtime Flag Injection via _configure Endpointtitle from the CVE record

Docker Model Runner (DMR) is software used to manage, run, and deploy AI models using Docker. Versions prior to 1.0.16 expose a POST `/engines/_configure` endpoint that accepts arbitrary runtime flags without authentication. These flags are passed directly to the underlying inference server (llama.cpp). By injecting the --log-file flag, an attacker with network access to the Model Runner API can write or overwrite arbitrary files accessible to the Model Runner process. When bundled with Docker Desktop (where Model Runner is enabled by default since version 4.46.0), it is reachable from any default container at model-runner.docker.internal without authentication. In this context, the file overwrite can target the Docker Desktop VM disk (`Docker.raw` ), resulting in the destruction of all containers, images, volumes, and build history. However, in specific configurations and with user interaction, it is possible to convert this vulnerability in a container escape. The issue is fixed in Docker Model Runner 1.0.16. Docker Desktop users should update to 4.61.0 or later, which includes the fixed Model Runner. A workaround is available. For Docker Desktop users, enabling Enhanced Container Isolation (ECI) blocks container access to Model Runner, preventing exploitation. However, if the Docker Model Runner is exposed to localhost over TCP in specific configurations, the vulnerability is still exploitable.

CVSS
7.5v3.1
EPSS (exploit probability)
0.3%18th pct
CISA KEV
Not listed

EPSS trendFIRST daily scores, last 90 days

FIRST EPSS held steady from 0.2% on 2026-10-19 to 0.3% on 2026-11-19, ranging 0.2%–0.3% across 32 daily scores.
2026-10-19 → 2026-11-19 · 32 days · 0.2%–0.3%
Daily values
FIRST EPSS daily scores for this CVE
Score dateEPSSPercentile
2026-11-190.3%17.7%
2026-11-180.3%17.6%
2026-11-170.3%17.6%
2026-11-160.3%17.6%
2026-11-150.3%17.6%
2026-11-140.3%17.6%
2026-11-130.3%17.6%
2026-11-120.3%17.6%
2026-11-110.3%17.6%
2026-11-100.3%17.3%
2026-11-090.3%17.3%
2026-11-080.3%17.3%
2026-11-070.3%17.3%
2026-11-060.3%17.3%
2026-11-050.3%17.3%
2026-11-040.3%17.3%
2026-11-030.3%17.3%
2026-11-020.3%17.3%
2026-11-010.3%17.3%
2026-10-310.3%17.3%
2026-10-300.3%17.3%
2026-10-290.3%17.2%
2026-10-280.3%17.2%
2026-10-270.3%17.1%
2026-10-260.3%17.1%
2026-10-250.3%17.1%
2026-10-240.2%6.3%
2026-10-230.2%6.3%
2026-10-220.2%6.3%
2026-10-210.2%6.3%
2026-10-200.2%6.3%
2026-10-190.2%6.3%

CVSS:3.1/AV:L/AC:H/PR:L/UI:R/S:C/C:H/I:H/A:H

Learn

Why it happened and what stops it

Coverage1 article

References

Facts sourced from NVD, FIRST EPSS, and CISA KEV. Last synced 2026-11-20 12:07 UTC.

  • CVSS & description: NVD 2.0, fetched 2026-11-18 18:07 UTC
  • EPSS: FIRST EPSS, fetched 2026-11-20 12:07 UTC; score date 2026-11-19
  • KEV: CISA KEV, fetched 2026-11-20 12:07 UTC