CVE-2026-5817
Docker Model Runner container-to-host code execution via unsandboxed trust_remote_code in Python inference backends
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered.
Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.
Attack profile
How easy it is to exploit
- Attack vector
- Local
- Complexity
- Low
- Privileges needed
- Low
- User interaction
- None
What an attacker can affect
- Confidentiality
- High
- Integrity
- High
- Availability
- High
Exploit likelihood
What each source reports
CVE List Score used
Docker Model Runner container-to-host code execution via unsandboxed trust_remote_code in Python inference backends
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered.
Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.
- Record
CVE-2026-5817- CVSS 4.0
- 8.8 High
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H - CVSS 3.1
- 8.2 High
CVSS:3.1/AV:L/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H - Weakness
- CWE-829
- Updated
- 2026-11-02
EUVDEUVD-2026-31493
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered.
Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.
- Record
EUVD-2026-31493- CVSS 4.0
- 8.8 High
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H - Updated
- 2026-11-02
GitHub AdvisoriesGHSA-cgqp-ww2v-6rjh
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets...
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered.
Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.
- Record
GHSA-cgqp-ww2v-6rjh- CVSS 4.0
- 8.8 High
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - CVSS 3.1
- 8.2 High
CVSS:3.1/AV:L/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H - Weakness
- CWE-829
- Updated
- 2026-11-02
NVD
The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered.
Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.
- Record
CVE-2026-5817- CVSS 4.0
- 8.8 High
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - CVSS 3.1
- 8.2 High
CVSS:3.1/AV:L/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H - CVSS 3.1
- 8.6 High
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H - Weakness
- CWE-829
- Updated
- 2026-11-02
Affected products
Public exploits
Links only. Every CVE stores no exploit code.
| Source | Link | Date |
|---|---|---|
| PoC in GitHub | davidrxchester/CVE-2026-5817 | 2026-05-20 |
| PoC in GitHub | gouldnicholas/CVE-2026-5817-PoC | 2026-05-22 |
History
| Recorded | Changed | Sources |
|---|---|---|
| first indexed | CVE List, EUVD, GitHub Advisories, NVD, PoC in GitHub |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
CVE-2026-5817 is a high vulnerability (CVSS 8.8) in Docker Desktop, published May 22, 2026. Public exploit code exists. EPSS puts the chance of exploitation in the next 30 days at 0.22%, the 13th percentile. No source names a fixed version.