| Aug 10 | Keras model loading is vulnerable to denial of service through HDF5 shape bombsCVE-2026-12570 KerasMedium5.5fixed in 3.15.0 | Keras | Medium5.5 | 3.15.0 |
| Aug 2 | Keras: HDF5 links can disclose local file contentsCVE-2026-9335 KerasMedium6.5fixed in 3.12.3, 3.15.0 | Keras | Medium6.5 | 3.12.3, 3.15.0 |
| Jul 19 | Keras: TorchModuleWrapper can deserialize unsafe PyTorch pickle dataCVE-2026-12484 KerasHigh7.8fixed in 3.12.3, 3.15.0 | Keras | High7.8 | 3.12.3, 3.15.0 |
| Jul 14 | Keras: tar extraction permits symlink-based path traversalCVE-2026-12482 KerasLow3.1fixed in 3.12.3, 3.15.0 | Keras | Low3.1 | 3.12.3, 3.15.0 |
| Jul 3 | Keras: Lambda deserialization can bypass safe mode and execute codeCVE-2026-12481 KerasHigh8.8fixed in 3.12.3, 3.15.0 | Keras | High8.8 | 3.12.3, 3.15.0 |
| Jul 1 | Keras: HDF5 virtual datasets can disclose local filesCVE-2026-12480 KerasMedium5.5fixed in 3.12.3, 3.15.0 | Keras | Medium5.5 | 3.12.3, 3.15.0 |
| Jun 22 | Keras: DiskIOStore permits path traversal through crafted layer namesCVE-2026-12479 KerasMedium6.1fixed in 3.12.3, 3.15.0 | Keras | Medium6.1 | 3.12.3, 3.15.0 |
| Jun 11 | Keras archive extraction utilities allow path traversal and arbitrary file writesCVE-2026-11816 KerasHigh8.1fixed in 3.14.0 | Keras | High8.1 | 3.14.0 |
| May 6 | Keras vulnerable to DoS via Malicious .keras Model (HDF5 Shape Bomb Causes Petabyte Allocation in KerasFileEditor)CVE-2026-0897 KerasHighfixed in 3.12.1, 3.13.2 | Keras | High | 3.12.1, 3.13.2 |
| Apr 13 | Keras has an untrusted deserialization vulnerabilityCVE-2026-1462 KerasHigh8.8fixed in 3.13.2 | Keras | High8.8 | 3.13.2 |
| Feb 18 | Keras has a Local File Disclosure via HDF5 External Storage During Keras Weight LoadingCVE-2026-1669 KerasHigh7.1fixed in 3.12.1, 3.13.2 | Keras | High7.1 | 3.12.1, 3.13.2 |
| Dec 22025 | Keras Directory Traversal VulnerabilityCVE-2025-12060 KerasHigh9.8fixed in 3.12.0 | Keras | High9.8 | 3.12.0 |
| Oct 292025 | Keras is vulnerable to arbitrary local file loading and Server-Side Request ForgeryCVE-2025-12058 KerasMediumfixed in 3.12.0 | Keras | Medium | 3.12.0 |
| Oct 172025 | Keras framework vulnerable to deserialization of untrusted dataCVE-2025-49655 KerasCritical9.8fixed in 3.11.3 | Keras | Critical9.8 | 3.11.3 |
| Sep 192025 | The Keras `Model.load_model` method **silently** ignores `safe_mode=True` and allows arbitrary code execution when a `.h5`/`.hdf5` file is loaded.CVE-2025-9905 KerasHighfixed in 3.11.3 | Keras | High | 3.11.3 |
| Sep 192025 | Keras is vulnerable to Deserialization of Untrusted DataCVE-2025-9906 KerasHigh7.3fixed in 3.11.0 | Keras | High7.3 | 3.11.0 |
| Aug 122025 | Keras vulnerable to CVE-2025-1550 bypass via reuse of internal functionalityCVE-2025-8747 KerasHigh8.8fixed in 3.11.0 | Keras | High8.8 | 3.11.0 |
| Mar 112025 | Arbitrary Code Execution via Crafted Keras Config for Model LoadingCVE-2025-1550 KerasHighfixed in 3.9.0 | Keras | High | 3.9.0 |
| Jan 82025 | keras Path Traversal vulnerabilityCVE-2024-55459 KerasMediumno fix yet | Keras | Medium | No fix yet |
| Jul 302024 | TensorFlow has segfault in array_ops.upper_boundCVE-2023-33976 TensorFlowHigh7.5fixed in 2.12.1 | TensorFlow | High7.5 | 2.12.1 |
| Apr 162024 | Keras code injection vulnerabilityCVE-2024-3660 KerasCritical9.8fixed in 2.13.1rc0 | Keras | Critical9.8 | 2.13.1rc0 |
| Mar 272023 | TensorFlow Denial of Service vulnerabilityCVE-2023-25661 TensorFlowMedium6.5fixed in 2.11.1 | TensorFlow | Medium6.5 | 2.11.1 |
| Mar 242023 | TensorFlow vulnerable to Out-of-Bounds Read in DynamicStitchCVE-2023-25659 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow vulnerable to seg fault in `tf.raw_ops.Print`CVE-2023-25660 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow vulnerable to integer overflow in EditDistanceCVE-2023-25662 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Null Pointer Error in TensorArrayConcatV2CVE-2023-25663 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Heap-buffer-overflow in AvgPoolGradCVE-2023-25664 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Null Pointer Error in SparseSparseMaximumCVE-2023-25665 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Floating Point Exception in AudioSpectrogramCVE-2023-25666 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow vulnerable to segfault when opening multiframe gifCVE-2023-25667 TensorFlowMedium6.5fixed in 2.11.1 | TensorFlow | Medium6.5 | 2.11.1 |
| Mar 242023 | TensorFlow has a heap out-of-buffer read vulnerability in the QuantizeAndDequantize operationCVE-2023-25668 TensorFlowCritical9.8fixed in 2.11.1 | TensorFlow | Critical9.8 | 2.11.1 |
| Mar 242023 | TensorFlow has Floating Point Exception in AvgPoolGrad with XLACVE-2023-25669 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Null Pointer Error in QuantizedMatMulWithBiasAndDequantizeCVE-2023-25670 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has segmentation fault in tfg-translateCVE-2023-25671 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Null Pointer Error in LookupTableImportV2CVE-2023-25672 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Floating Point Exception in TensorListSplit with XLACVE-2023-25673 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Null Pointer Error in RandomShuffle with XLA enableCVE-2023-25674 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has Segfault in Bincount with XLACVE-2023-25675 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has null dereference on ParallelConcat with XLACVE-2023-25676 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow has double free in Fractional(Max/Avg)PoolCVE-2023-25801 TensorFlowHigh8.0fixed in 2.11.1 | TensorFlow | High8.0 | 2.11.1 |
| Mar 242023 | TensorFlow has Floating Point Exception in TFLite in conv kernelCVE-2023-27579 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Mar 242023 | TensorFlow vulnerable to Out-of-Bounds Read in GRUBlockCellGradCVE-2023-25658 TensorFlowHigh7.5fixed in 2.11.1 | TensorFlow | High7.5 | 2.11.1 |
| Nov 222022 | Tensorflow vulnerable to Out-of-Bounds ReadCVE-2022-41880 TensorFlowMedium6.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium6.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK` failure in `SobolSample` via missing validationGHSA-cqvq-fvhr-v6hc TensorFlowLowfixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Low | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK` fail in `TensorListScatter` and `TensorListScatterV2` in eager modeGHSA-xf83-q765-xm6m TensorFlowLowfixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Low | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Invalid char to bool conversion when printing a tensorCVE-2022-41911 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Heap overflow in `QuantizeAndDequantizeV2`CVE-2022-41910 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Segfault in `CompositeTensorVariantToComponents`CVE-2022-41909 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK` fail via inputs in `PyFunc`CVE-2022-41908 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Overflow in `ResizeNearestNeighborGrad`CVE-2022-41907 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Out of bounds write in grappler in TensorflowCVE-2022-41902 TensorFlowHigh7.1fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | High7.1 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK_EQ` fail via input in `SparseMatrixNNZ`CVE-2022-41901 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | FractionalMaxPool and FractionalAVGPool heap out-of-bounds acessCVE-2022-41900 TensorFlowHigh7.1fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | High7.1 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK` fail via inputs in `SdcaOptimizer`CVE-2022-41899 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK` fail via inputs in `SparseFillEmptyRowsGrad`CVE-2022-41898 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `FractionalMaxPoolGrad` Heap out of bounds readCVE-2022-41897 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `tf.raw_ops.Mfcc` crashesCVE-2022-41896 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `MirrorPadGrad` heap out of bounds readCVE-2022-41895 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | Buffer overflow in `CONV_3D_TRANSPOSE` on TFLiteCVE-2022-41894 TensorFlowHigh7.1fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | High7.1 | 2.8.4, 2.9.3, 2.10.1 |
| Nov 212022 | `CHECK_EQ` fail in `tf.raw_ops.TensorListResize`CVE-2022-41893 TensorFlowMedium4.8fixed in 2.8.4, 2.9.3, 2.10.1 | TensorFlow | Medium4.8 | 2.8.4, 2.9.3, 2.10.1 |