StackPrism

PyPI package

torch 2.7.0 vulnerabilities, and what to upgrade to

torch 2.7.0 has 14 known vulnerabilities; 2.13.0 fixes 13 of them, and 1 has no fixed version yet.

PyPI (opens in a new tab) describes it as “Tensors and Dynamic neural networks in Python with strong GPU acceleration”

Known vulnerabilities
14
None rated high or critical
Released
Apr 23, 2025
Latest is 2.14.1
Licence
Apache-2.0 AND Apache-2.0 WITH LLVM-exception AND BSD-2-Clause AND BSD-3-Clause AND BSL-1.0 AND MIT
Pulls in
24 packages
Resolved by deps.dev

What to upgrade to

2.13.0 is the closest release with no known vulnerabilities.

  • 2.13.0Upgrade to this

    Fixes 13 of the 14 known vulnerabilities. Released Jul 8, 2026.

    No known vulnerabilities

  • 2.14.1

    Newest 2.x release and the latest release. Released Sep 30, 2026.

    No known vulnerabilities

2.13.0 came out on Jul 8, 2026. A coding agent trained before then doesn’t know it exists, so it may suggest an older release. AI knowledge cutoff dates

Hand the fix to your coding agent

One task with the upgrade, the file to change and why. Paste it into Claude Code, Cursor or Codex.

Findings

Every advisory for torch 2.7.0

Known exploited first (CISA KEV), then by how likely they are to be exploited (FIRST EPSS), then by severity. Each id links to its OSV.dev record.

  1. WarningCVE-2026-24747 on OSV.dev (opens in a new tab)GHSA-63cw-57p8-fm3pEPSS: top 44%

    PyTorch is a Python package that provides tensor computation. Prior to version 2.10.0, a vulnerability in PyTorch's `weights_only` unpickler allows an attacker to craft a malicious checkpoint file (`…

    Fixed in 2.10.0. Published Jan 27, 2026.

  2. WarningCVE-2025-55558 on OSV.dev (opens in a new tab)PYSEC-2025-208EPSS: top 63%

    A buffer overflow occurs in pytorch v2.7.0 when a PyTorch model consists of torch.nn.Conv2d, torch.nn.functional.hardshrink, and torch.Tensor.view-torch.mv() and is compiled by Inductor, leading to a…

    Fixed in 2.7.1. Published Sep 25, 2025.

  3. WarningCVE-2025-55551 on OSV.dev (opens in a new tab)PYSEC-2025-203EPSS: top 65%

    An issue in the component torch.linalg.lu of pytorch v2.8.0 allows attackers to cause a Denial of Service (DoS) when performing a slice operation.

    Fixed in 2.9.0. Published Sep 25, 2025.

  4. WarningCVE-2025-55552 on OSV.dev (opens in a new tab)PYSEC-2025-204EPSS: top 66%

    pytorch v2.8.0 was discovered to display unexpected behavior when the components torch.rot90 and torch.randn_like are used together.

    Fixed in 2.9.0. Published Sep 25, 2025.

  5. WarningCVE-2025-55560 on OSV.dev (opens in a new tab)PYSEC-2025-209EPSS: top 67%

    An issue in pytorch v2.7.0 can lead to a Denial of Service (DoS) when a PyTorch model consists of torch.Tensor.to_sparse() and torch.Tensor.to_dense() and is compiled by Inductor.

    Fixed in 2.7.1. Published Sep 25, 2025.

  6. WarningCVE-2025-55557 on OSV.dev (opens in a new tab)PYSEC-2025-207EPSS: top 67%

    A Name Error occurs in pytorch v2.7.0 when a PyTorch model consists of torch.cummin and is compiled by Inductor, leading to a Denial of Service (DoS).

    Fixed in 2.7.1. Published Sep 25, 2025.

  7. WarningCVE-2025-55553 on OSV.dev (opens in a new tab)PYSEC-2025-205EPSS: top 67%

    A syntax error in the component proxy_tensor.py of pytorch v2.7.0 allows attackers to cause a Denial of Service (DoS).

    Fixed in 2.7.1. Published Sep 25, 2025.

  8. WarningCVE-2026-4538 on OSV.dev (opens in a new tab)PYSEC-2026-139EPSS: top 68%

    A vulnerability was identified in PyTorch 2.10.0. The affected element is an unknown function of the component pt2 Loading Handler. The manipulation leads to deserialization. The attack can only be p…

    No fixed version listed yet. Published Mar 22, 2026.

  9. WarningCVE-2025-3730 on OSV.dev (opens in a new tab)GHSA-887c-mr87-cxwpEPSS: top 76%

    PyTorch Improper Resource Shutdown or Release vulnerability

    Fixed in 2.8.0. Published Apr 16, 2025.

  10. WarningCVE-2025-55554 on OSV.dev (opens in a new tab)PYSEC-2025-206EPSS: top 78%

    pytorch v2.8.0 was discovered to contain an integer overflow in the component torch.nan_to_num-.long().

    Fixed in 2.9.0. Published Sep 25, 2025.

  11. LowCVE-2025-2953 on OSV.dev (opens in a new tab)GHSA-3749-ghw9-m3mgEPSS: top 84%

    PyTorch susceptible to local Denial of Service

    Fixed in 2.7.1-rc1. Published Mar 30, 2025.

  12. LowCVE-2025-3000 on OSV.dev (opens in a new tab)GHSA-rrmf-rvhw-rf47EPSS: top 91%

    PyTorch is vulnerable to memory corruption through its torch.jit.script function

    Fixed in 2.13.0. Published Mar 31, 2025.

  13. WarningCVE-2025-2999 on OSV.dev (opens in a new tab)GHSA-vgrw-7cvw-pwgxEPSS: top 92%

    PyTorch is vulnerable to memory corruption through its unpack_sequence function

    Fixed in 2.9.1. Published Mar 31, 2025.

  14. LowCVE-2025-3001 on OSV.dev (opens in a new tab)GHSA-qfhq-4f3w-5fphEPSS: top 92%

    PyTorch is vulnerable to memory corruption through its torch.lstm_cell function

    Fixed in 2.10.0. Published Mar 31, 2025.

Is torch the only one?

Paste a public GitHub repo or drop your requirements.txt or pyproject.toml and see every vulnerable, end-of-life or deprecated dependency, with one fix task. Free, no sign-up.

Check your whole repo

How this page is checked

StackPrism asks OSV.dev which advisories affect exactly torch 2.7.0, merges duplicates filed under several ids, and ranks them with CISA’s Known Exploited Vulnerabilities list and FIRST EPSS. Release dates and deprecation come from deps.dev and PyPI. No AI writes this page: every sentence comes from that data, refreshed every few hours.

Last checked against PyPI, OSV.dev, deps.dev, CISA KEV and FIRST EPSS.