UTC
High

PyTorch Lightning load_from_checkpoint has an insecure checkpoint deserialization

CVE-2026-31221 · GHSA-75m9-98v2-hjpm · published 2026-05-12 · updated 2026-05-18

Severity
High
CVSS score
7.8
Chance of exploitation (EPSS)
0.39%
Fixed in
No fix yet
  • Exploitation No known exploitation
  • Automatable No
  • Technical impact Total
  • CISA SSVC
What it is

PyTorch-Lightning versions 2.6.0 and earlier contain an insecure deserialization vulnerability (CWE-502) in the checkpoint loading mechanism. The LightningModule.load_from_checkpoint() method, which is commonly used to load saved model states, internally calls torch.load() without setting the security-restrictive weights_only=True parameter. This default behavior allows the deserialization of arbitrary Python objects via the Pickle module. A remote attacker can exploit this by providing a maliciously crafted checkpoint file, leading to arbitrary code execution on the victim's system when the file is loaded.

What to do

No fixed release is listed yet for pytorch-lightning. Follow the source advisory for updates, and until one ships:

General guidance from ShieldPi, not a vendor fix

  • Unsafe deserialization. Do not let it load files, models or data from sources you do not trust, and keep its RPC and socket ports reachable only from localhost or hosts you trust.
Affected
PackageRegistryAffectedFixed in
pytorch-lightningPyPI<= 2.6.0No fix yet
Also in pytorch-lightning
  1. CriticalCompromise of PyTorch Lightning PyPi Package Versionspytorch-lightning, pytorch-lightning
  2. CriticalPyTorch Lightning path traversal vulnerabilitylightning, pytorch-lightning
  3. Criticalpytorch-lightning vulnerable to Arbitrary File Write via /v1/runs API endpointlightning, pytorch-lightning
  4. CriticalRemote code execution in pytorch lightninglightning, pytorch-lightning
  5. CriticalCode Injection in PyTorch Lightningpytorch-lightning
  6. HighPyTorch Lightning denial of service vulnerabilitypytorch-lightning

The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].