CVE-2026-58659: PyTorch Lightning allows arbitrary code execution through checkpoint _instantiator hyperparameters
(updated )
PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the _load_state function that imports and executes attacker-controlled module names from checkpoint _instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass weights_only=True protections to execute arbitrary code when LightningModule.load_from_checkpoint is called.
References
- github.com/Lightning-AI/pytorch-lightning/commit/d710d689510d50e800f53b3cd773cbca20b1f86f
- github.com/Lightning-AI/pytorch-lightning/issues/21822
- github.com/Lightning-AI/pytorch-lightning/pull/21832
- github.com/Lightning-AI/pytorch-lightning/releases/tag/2.6.6
- github.com/advisories/GHSA-qqmf-gpg7-g8gw
- github.com/pypa/advisory-database/tree/main/vulns/lightning/PYSEC-2026-3624.yaml
- github.com/pypa/advisory-database/tree/main/vulns/pytorch-lightning/PYSEC-2026-3967.yaml
- nvd.nist.gov/vuln/detail/CVE-2026-58659
- www.vulncheck.com/advisories/pytorch-lightning-arbitrary-code-execution-via-instantiator-hyperparameter
Code Behaviors & Features
Detect and mitigate CVE-2026-58659 with GitLab Dependency Scanning
Secure your software supply chain by verifying that all open source dependencies used in your projects contain no disclosed vulnerabilities. Dependency Scanning using SBOM also detects vulnerabilities from proprietary advisories that this site does not include. Learn more about Dependency Scanning using SBOM →