CVE-2023-25676: TensorFlow GPU Null Dereference Vulnerability in ParallelConcat
Summary
CVE-2023-25676: TensorFlow has null dereference on ParallelConcat with XLA. Affects tensorflow-gpu versions < 2.11.1. Fixed in 2.11.1. How to fix ↓
Also tracked for
- tensorflow · PyPI
- tensorflow-cpu · PyPI
- tensorflow · Bitnami
- tensorflow · PyPI
The same CVE-2023-25676 is tracked separately for each ecosystem and vendor. Its fix information is the same; the package and affected versions differ.
Check if this vulnerability affects your application
Upload a manifest or SBOM and we'll show whether you're affected, plus every other risky package in the same scan.
No signup required · Results in about 60 seconds
Background
TensorFlow is an open-source machine learning framework widely used for building and training machine learning models. It provides a comprehensive ecosystem for developing deep learning applications, including tools for data manipulation, model training, and deployment. The TensorFlow GPU package enhances performance by utilizing GPU resources for computation. However, it is susceptible to vulnerabilities that can lead to application crashes or unexpected behavior, particularly when using specific operations like ParallelConcat with XLA.
Vulnerability Detail
The vulnerability arises in TensorFlow's raw operations, specifically in the tf.raw_ops.ParallelConcat function when used with XLA (Accelerated Linear Algebra). If the 'shape' parameter is provided with a rank that is not greater than zero, it leads to a null pointer dereference, causing a segmentation fault (segfault). This can result in application crashes, potentially disrupting services that rely on TensorFlow for machine learning tasks. The vulnerability is critical as it can be exploited easily during model execution, affecting the stability of applications.
How To Fix
Latest published version: 2.12.0 as of 22 Oct 2026
pip install --upgrade "tensorflow-gpu==2.12.0"Upgrading to it may include breaking changes; the version below is the minimum fixed release for this vulnerability.
If you need the minimum change
2.11.1
This is the first release that contains the fix.
pip install --upgrade "tensorflow-gpu==2.11.1"Tip — Not sure which version your app runs? Vulert can match this fix to it. Scan free →
Workaround
As a temporary workaround, avoid using the tf.raw_ops.ParallelConcat function with a 'shape' parameter that has a rank of zero or less. For example, ensure that the shape parameter is defined correctly as follows:
import tensorflow as tf
func = tf.raw_ops.ParallelConcat
para = {'shape': [1], 'values': [1]} # Ensure shape is greater than zero
@tf.function(jit_compile=True)
def test():
y = func(**para)
return y
test()
Conclusion
In summary, CVE-2023-25676 is a critical vulnerability in TensorFlow GPU that can lead to application crashes due to a null dereference in the ParallelConcat function. It is essential to update to version 2.11.1 or later to fix this issue. Users should also consider using Vulert for real-time monitoring of vulnerabilities in their applications.
See every vulnerability recorded for tensorflow-gpu →
FAQ
CVE-2023-25676 is a vulnerability in TensorFlow GPU that causes a null dereference in the ParallelConcat function when used with an invalid shape parameter.
This vulnerability can lead to application crashes, especially if your application relies on TensorFlow for machine learning tasks and uses the ParallelConcat operation incorrectly.
The vulnerability affects TensorFlow GPU versions prior to 2.11.1. It is recommended to upgrade to this version or later to mitigate the risk.
You can use Vulert to monitor and alert for open-source vulnerabilities in your software. Visit the Vulert playground at Vulert Scanner to check if your application is affected without any signup.
If an immediate upgrade is not possible, ensure that you do not use the ParallelConcat function with a shape parameter that has a rank of zero or less as a temporary workaround.
References
Related vulnerabilities
Recent vulnerabilities recorded for tensorflow-gpu.
