PYSEC-2026-3624
Vulnerability from pysec - Published: 2026-07-15 18:16 - Updated: 2026-08-07 09:08
VLAI
Details
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.
Severity
Impacted products
| Name | purl | lightning | pkg:pypi/lightning |
|---|
Aliases
{
"affected": [
{
"ecosystem_specific": {},
"package": {
"ecosystem": "PyPI",
"name": "lightning",
"purl": "pkg:pypi/lightning"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2022.6.15"
}
],
"type": "ECOSYSTEM"
}
],
"versions": [
"1.8.0",
"1.8.0.post1",
"1.8.0rc1",
"1.8.0rc2",
"1.8.1",
"1.8.2",
"1.8.3",
"1.8.3.post0",
"1.8.3.post1",
"1.8.3.post2",
"1.8.4",
"1.8.4.post0",
"1.8.5",
"1.8.5.post0",
"1.8.6",
"1.9.0",
"1.9.0rc0",
"1.9.1",
"1.9.2",
"1.9.3",
"1.9.4",
"1.9.5",
"2.0.0",
"2.0.0rc0",
"2.0.1",
"2.0.1.post0",
"2.0.2",
"2.0.3",
"2.0.4",
"2.0.5",
"2.0.6",
"2.0.7",
"2.0.8",
"2.0.9",
"2.0.9.post0",
"2.1.0",
"2.1.0rc0",
"2.1.0rc1",
"2.1.1",
"2.1.2",
"2.1.3",
"2.1.4",
"2.2.0",
"2.2.0.post0",
"2.2.0rc0",
"2.2.1",
"2.2.2",
"2.2.3",
"2.2.4",
"2.2.5",
"2.3.0",
"2.3.0.dev20240318",
"2.3.0.dev20240324",
"2.3.0.dev20240328",
"2.3.0.dev20240331",
"2.3.0.dev20240407",
"2.3.0.dev20240414",
"2.3.0.dev20240421",
"2.3.0.dev20240428",
"2.3.0.dev20240505",
"2.3.0.dev20240519",
"2.3.0.dev20240526",
"2.3.0.dev20240602",
"2.3.0.dev20240609",
"2.3.0.dev20240616",
"2.3.0.dev20240623",
"2.3.1",
"2.3.2",
"2.3.3",
"2.4.0",
"2.4.0.dev20240630",
"2.4.0.dev20240707",
"2.4.0.dev20240714",
"2.4.0.dev20240721",
"2.4.0.dev20240728",
"2.4.0.dev20240804",
"2.4.0.dev20240811",
"2.4.0.dev20240818",
"2.4.0.dev20240825",
"2.4.0.dev20240901",
"2.4.0.dev20240908",
"2.4.0.dev20240915",
"2.4.0.dev20240922",
"2.4.0.dev20240929",
"2.4.0.dev20241006",
"2.4.0.dev20241013",
"2.4.0.dev20241020",
"2.5.0",
"2.5.0.dev20241027",
"2.5.0.dev20241103",
"2.5.0.dev20241110",
"2.5.0.dev20241117",
"2.5.0.dev20241124",
"2.5.0.dev20241201",
"2.5.0.dev20241208",
"2.5.0.dev20241215",
"2.5.0.dev20241222",
"2.5.0.dev20241229",
"2.5.0.dev20250105",
"2.5.0.dev20250112",
"2.5.0.dev20250119",
"2.5.0.dev20250126",
"2.5.0.dev20250202",
"2.5.0.dev20250209",
"2.5.0.dev20250216",
"2.5.0.dev20250223",
"2.5.0.dev20250302",
"2.5.0.dev20250309",
"2.5.0.dev20250316",
"2.5.0.post0",
"2.5.0rc0",
"2.5.1",
"2.5.1.dev20250323",
"2.5.1.dev20250330",
"2.5.1.dev20250406",
"2.5.1.dev20250413",
"2.5.1.dev20250420",
"2.5.1.dev20250427",
"2.5.1.dev20250504",
"2.5.1.dev20250511",
"2.5.1.dev20250518",
"2.5.1.dev20250525",
"2.5.1.dev20250601",
"2.5.1.dev20250608",
"2.5.1.dev20250615",
"2.5.1.dev20250622",
"2.5.1.post0",
"2.5.1rc0",
"2.5.1rc1",
"2.5.1rc2",
"2.5.2",
"2.5.3",
"2.5.4",
"2.5.5",
"2.5.6",
"2.6.0",
"2.6.0.dev0",
"2.6.0.dev20250629",
"2.6.0.dev20250706",
"2.6.0.dev20250713",
"2.6.0.dev20250720",
"2.6.0.dev20250727",
"2.6.0.dev20250803",
"2.6.0.dev20250810",
"2.6.0.dev20250817",
"2.6.0.dev20250824",
"2.6.0.dev20250831",
"2.6.0.dev20250914",
"2.6.0.dev20250921",
"2.6.0.dev20250928",
"2.6.0.dev20251005",
"2.6.0.dev20251012",
"2.6.0.dev20251019",
"2.6.0.dev20251026",
"2.6.0.dev20251102",
"2.6.0.dev20251109",
"2.6.0.dev20251116",
"2.6.0.dev20251123",
"2.6.0.dev20251130",
"2.6.0.dev20251207",
"2.6.0.dev20251214",
"2.6.0.dev20251221",
"2.6.0.dev20251228",
"2.6.0.dev20260104",
"2.6.0.dev20260111",
"2.6.0.dev20260118",
"2.6.0.dev20260125",
"2.6.1",
"2.6.1.dev20260201",
"2.6.4",
"2.6.5"
]
}
],
"aliases": [
"CVE-2026-58659"
],
"details": "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.",
"id": "PYSEC-2026-3624",
"modified": "2026-08-07T09:08:04.834348Z",
"published": "2026-07-15T18:16:48.743Z",
"references": [
{
"type": "REPORT",
"url": "https://github.com/Lightning-AI/pytorch-lightning/issues/21822"
},
{
"type": "FIX",
"url": "https://github.com/Lightning-AI/pytorch-lightning/commit/d710d689510d50e800f53b3cd773cbca20b1f86f"
},
{
"type": "FIX",
"url": "https://github.com/Lightning-AI/pytorch-lightning/pull/21832"
},
{
"type": "FIX",
"url": "https://www.vulncheck.com/advisories/pytorch-lightning-arbitrary-code-execution-via-instantiator-hyperparameter"
}
],
"severity": [
{
"score": "CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/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",
"type": "CVSS_V4"
}
]
}
Loading…
Loading…
Experimental. This forecast is provided for visualization only and may change without notice. Do not use it for operational decisions.
Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Loading…
The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the
vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base
model, served locally by ML-Gateway.
They have not been verified by an analyst and are provided for guidance only.
Loading…
Loading…