PYSEC-2026-3084 Stack overflow in TensorFlow

Impact The GraphDef format in TensorFlow does not allow self recursive functions. The runtime assumes that this invariant is satisfied. However, a GraphDef containing a fragment such as the following can be consumed when loading a SavedModel: library function signature name: "SomeOp" description:...

PYSEC-2026-3225 Crash due to erroneous `StatusOr` in TensorFlow

Impact A GraphDef from a TensorFlow SavedModel can be maliciously altered to cause a TensorFlow process to crash due to encountering a StatusOr value that is an error and forcibly extracting the value from it: cc if opregdata-typector != nullptr VLOG3 opdef; const FullTypeDef ctortypedef =...

Out of bounds read in Tensorflow

ImpactThe TFG dialect of TensorFlow MLIR makes several assumptions about the incoming GraphDef before converting it to the MLIR-based dialect.If an attacker changes the SavedModel format on disk to invalidate these assumptions and the GraphDef is then converted to MLIR-based IR then they can caus...

PYSEC-2026-3137 Integer overflow in TensorFlow

Impact Under certain scenarios, Grappler component of TensorFlow is vulnerable to an integer overflow during cost estimation for crop and resize. Since the cropping parameters are user controlled, a malicious person can trigger undefined behavior. Patches We have patched the issue in GitHub commi...

Out of bounds read in Tensorflow

Impact The implementation of FractionalAvgPoolGrad does not consider cases where the input tensors are invalid allowing an attacker to read from outside of bounds of heap:pythonimport tensorflow as [email protected] test: y = tf.rawops.FractionalAvgPoolGrad originputtensorshape=2,2,2,2,...

Uninitialized variable access in Tensorflow

ImpactThe implementation of AssignOp can result in copying unitialized data to a new tensor. This later results in undefined behavior.The implementation has a check that the left hand side of the assignment is initialized to minimize number of allocations, but does not check that the right hand...

Memory leak in decoding PNG images

ImpactWhen decoding PNG images TensorFlow can produce a memory leak if the image is invalid.After calling png::CommonInitDecode..., , the decode value contains allocated buffers which can only be freed by calling png::CommonFreeDecode. However, several error case in the function implementation...

Integer overflow in TensorFlow

ImpactUnder certain scenarios, Grappler component of TensorFlow is vulnerable to an integer overflow during cost estimation for crop and resize. Since the cropping parameters are user controlled, a malicious person can trigger undefined behavior. PatchesWe have patched the issue in GitHub commit...

Out of bounds read and write in Tensorflow

ImpactThere is a typo in TensorFlow's SpecializeType which results in heap OOB read/write:ccfor int i = 0; i argssize; j++ auto arg = t-mutableargsi; // ... Due to a typo, arg is initialized to the ith mutable argument in a loop where the loop index is j. Hence it is possible to assign to arg fro...

PYSEC-2026-3247 Out of bounds read in Tensorflow

Impact The implementation of FractionalAvgPoolGrad does not consider cases where the input tensors are invalid allowing an attacker to read from outside of bounds of heap: python import tensorflow as tf @tf.function def test: y = tf.rawops.FractionalAvgPoolGrad originputtensorshape=2,2,2,2,...

PYSEC-2026-3096 Multiple `CHECK`-fails in `function.cc` in TensowFlow

Impact A malicious user can cause a denial of service by altering a SavedModel such that assertions in function.cc would be falsified and crash the Python interpreter. Patches We have patched the issue in GitHub commits dcc21c7bc972b10b6fb95c2fb0f4ab5a59680ec2 and...

PYSEC-2026-3181 `CHECK`-fails due to attempting to build a reference tensor

Impact A malicious user can cause a denial of service by altering a SavedModel such that Grappler optimizer would attempt to build a tensor using a reference dtype. This would result in a crash due to a CHECK-fail in the Tensor constructor as reference types are not allowed. Patches We have patch...

Out of bounds read in Tensorflow

Impact The implementation of shape inference for ReverseSequence does not fully validate the value of batchdim and can result in a heap OOB read:pythonimport tensorflow as [email protected] test: y = tf.rawops.ReverseSequence input = 'aaa','bbb', seqlengths = 1,1,1, seqdim = -10, batchdim = -10...

Multiple `CHECK`-fails in `function.cc` in TensowFlow

ImpactA malicious user can cause a denial of service by altering a SavedModel such that assertions in function.cc would be falsified and crash the Python interpreter. PatchesWe have patched the issue in GitHub commits dcc21c7bc972b10b6fb95c2fb0f4ab5a59680ec2 and...

PYSEC-2026-3229 Uninitialized variable access in Tensorflow

Impact The implementation of AssignOp can result in copying unitialized data to a new tensor. This later results in undefined behavior. The implementation has a check that the left hand side of the assignment is initialized to minimize number of allocations, but does not check that the right hand...

PYSEC-2026-3113 Abort caused by allocating a vector that is too large in Tensorflow

Impact During shape inference, TensorFlow can allocate a large vector based on a value from a tensor controlled by the user: cc const auto numdims = Valueshapedim; std::vector dims; dims.reservenumdims; Patches We have patched the issue in GitHub commit 1361fb7e29449629e1df94d44e0427ebec8c83c7. T...

Out of bounds read in Tensorflow

Impact The implementation of Dequantize does not fully validate the value of axis and can result in heap OOB accesses:pythonimport tensorflow as [email protected] test: y = tf.rawops.Dequantize input=tf.constant1,1,dtype=tf.qint32, minrange=1.0, maxrange=10.0, mode='MINCOMBINED', narrowrange=Fals...

Integer overflow in Tensorflow

Impact The implementation of shape inference for Dequantize is vulnerable to an integer overflow weakness:pythonimport tensorflow as tfinput = tf.constant1,1,[email protected] test: y = tf.rawops.Dequantize input=input, minrange=1.0, maxrange=10.0, mode='MINCOMBINED',...

Abort caused by allocating a vector that is too large in Tensorflow

ImpactDuring shape inference, TensorFlow can allocate a large vector based on a value from a tensor controlled by the user:cc const auto numdims = Valueshapedim; std::vector dims; dims.reservenumdims; Patches We have patched the issue in GitHub commit 1361fb7e29449629e1df94d44e0427ebec8c83c7.The...

`CHECK`-failures during Grappler's `IsSimplifiableReshape` in Tensorflow

ImpactThe Grappler optimizer in TensorFlow can be used to cause a denial of service by altering a SavedModel such that IsSimplifiableReshape would trigger CHECK failures. PatchesWe have patched the issue in GitHub commits ebc1a2ffe5a7573d905e99bd0ee3568ee07c12c1,...