PT-2026-59991

Impact DenseBincount assumes its input tensor weights to either have the same shape as its input tensor input or to be length-0. A different weights shape will trigger a CHECK fail that can be used to trigger a denial of service attack. python import tensorflow as tf binary output = True input =...

PT-2026-59917

Impact If LRNGrad is given an output image input tensor that is not 4-D, it results in a CHECK fail that can be used to trigger a denial of service attack. python import tensorflow as tf depth radius = 1 bias = 1.59018219 alpha = 0.117728651 beta = 0.404427052 input grads =...

PT-2026-59900

Impact If QuantizedMatMul is given nonscalar input for: - min a - max a - min b - max b It gives a segfault that can be used to trigger a denial of service attack. python import tensorflow as tf Toutput = tf.qint32 transpose a = False transpose b = False Tactivation = tf.quint8 a = tf.constant7,...

PT-2026-59737

Impact If QuantizedMatMul is given nonscalar input for: - min a - max a - min b - max b It gives a segfault that can be used to trigger a denial of service attack. python import tensorflow as tf Toutput = tf.qint32 transpose a = False transpose b = False Tactivation = tf.quint8 a = tf.constant7,...

PT-2026-59894

Impact Constructing a tflite model with a paramater filter input channel of less than 1 gives a FPE. Patches We have patched the issue in GitHub commit 34f8368c535253f5c9cb3a303297743b62442aaa. The fix will be included in TensorFlow 2.12. We will also cherrypick this commit on TensorFlow 2.11.1...

PT-2026-59644

Impact An attacker who uses this vulnerability can craft a PDF which leads to an infinite loop. This requires merging a file with outlines into a writer. Patches This has been fixed in pypdf==6.13.0. Workarounds If you cannot upgrade yet, consider applying the changes from PR 3830...

PT-2026-59722

Impact If QuantizedAvgPool is given min input or max input tensors of a nonzero rank, it results in a segfault that can be used to trigger a denial of service attack. python import tensorflow as tf ksize = 1, 2, 2, 1 strides = 1, 2, 2, 1 padding = "SAME" input = tf.constant1, shape=1,4,4,2,...

PT-2026-59942

Impact An input sparse matrix that is not a matrix with a shape with rank 0 will trigger a CHECK fail in tf.raw ops.SparseMatrixNNZ. python import tensorflow as tf tf.raw ops.SparseMatrixNNZsparse matrix= Patches We have patched the issue in GitHub commit f856d02e5322821aad155dad9b3acab1e9f5d693...

PT-2026-59876

Impact An input pooling ratio that is smaller than 1 will trigger a heap OOB in tf.raw ops.FractionalMaxPool and tf.raw ops.FractionalAvgPool. Patches We have patched the issue in GitHub commit 216525144ee7c910296f5b05d214ca1327c9ce48. The fix will be included in TensorFlow 2.11.0. We will also...

PT-2026-59665

Impact A Python operator precedence bug in pyzipper/zipfile aes.py caused the AE-2 format to never be automatically selected during encryption, regardless of file size or compression type. As a result, all encrypted entries are written in AE-1 format unless AE-2 is explicitly forced by the caller...

PT-2026-59747

Impact TFversion 2.11.0 //tensorflow/core/ops/array ops.cc:1067 const Tensor hypothesis shape t = c-input tensor2; std::vector dimshypothesis shape t-NumElements - 1; for int i = 0; i MakeDimstd::maxh valuesi, t valuesi; if hypothesis shape t is empty, hypothesis shape t-NumElements - 1 will be...

PT-2026-59733

Impact FPE in TensorListSplit with XLA python import tensorflow as tf func = tf.raw ops.TensorListSplit para = 'tensor': 1, 'element shape': -1, 'lengths': 0 @tf.functionjit compile=True def fuzz jit: y = funcpara return y printfuzz jit Patches We have patched the issue in GitHub commit...

PT-2026-59736

Impact If tf.raw ops.TensorListResize is given a nonscalar value for input size, it results CHECK fail which can be used to trigger a denial of service attack. python import numpy as np import tensorflow as tf a = data structures.tf tensor list newelements = tf.constantvalue=3, 4, 5 b = np.zeros0...

PT-2026-59163

Predictable bucket naming in Vertex AI Experiments in Google Cloud Vertex AI from version 1.21.0 up to but not including 1.133.0 on Google Cloud Platform allows an unauthenticated remote attacker to achieve cross-tenant remote code execution, model theft, and poisoning via pre-creating predictabl...

PT-2026-59586

A vulnerability was found in PrefectHQ prefect up to 3.6.25.dev6. Affected by this issue is some unknown functionality of the file src/prefect/runner/storage.py of the component GitRepository Pull Handler. The manipulation of the argument commit sha/directories results in argument injection. It i...

PT-2026-59613

Impact OGC API - Process execution requests can use the subscriber object to requests to internal HTTP services. Patches The issue has been patched in master branch and made available as part of the 0.23.3 release. The patch disables any HTTP requests made to internal resources by default unless...

PT-2026-60008

Summary The utcp-http plugin is vulnerable to a blind Server-Side Request Forgery SSRF caused by a trust-boundary inconsistency between manual discovery and tool invocation. register manual validates the discovery URL against an HTTPS / loopback allowlist, but call tool and call tool streaming...

PT-2026-59993

Impact If Requantize is given input min, input max, requested output min, requested output max tensors of a nonzero rank, it results in a segfault that can be used to trigger a denial of service attack. python import tensorflow as tf out type = tf.quint8 input = tf.constant1, shape=3,...

PT-2026-59928

Impact nn ops.fractional avg pool v2 and nn ops.fractional max pool v2 require the first and fourth elements of their parameter pooling ratio to be equal to 1.0, as pooling on batch and channel dimensions is not supported. python import tensorflow as tf import os import numpy as np from...

PT-2026-59799

Impact A malicious invalid input crashes a tensorflow model Check Failed and can be used to trigger a denial of service attack. To minimize the bug, we built a simple single-layer TensorFlow model containing a Convolution3DTranspose layer, which works well with expected inputs and can be deployed...