CVE-2023-25672
📋 TL;DR
This vulnerability in TensorFlow's LookupTableImportV2 function causes a Null Pointer Exception (NPE) when scalar values are passed, potentially leading to denial of service. It affects TensorFlow users who utilize this specific function in their machine learning workflows. The vulnerability could crash TensorFlow processes when exploited.
💻 Affected Systems
- TensorFlow
📦 What is this software?
⚠️ Risk & Real-World Impact
Worst Case
Complete denial of service causing TensorFlow applications to crash, disrupting machine learning inference or training pipelines.
Likely Case
Application crashes or unexpected termination when LookupTableImportV2 is called with scalar inputs, requiring process restart.
If Mitigated
Minimal impact if proper input validation is implemented or the vulnerable function is not used.
🎯 Exploit Status
Exploitation requires calling the vulnerable function with scalar inputs, which may require specific application logic.
🛠️ Fix & Mitigation
✅ Official Fix
Patch Version: TensorFlow 2.12.0 or 2.11.1
Vendor Advisory: https://github.com/tensorflow/tensorflow/security/advisories/GHSA-94mm-g2mv-8p7r
Restart Required: Yes
Instructions:
1. Update TensorFlow using pip: 'pip install --upgrade tensorflow==2.12.0' or 'pip install --upgrade tensorflow==2.11.1'. 2. Restart all TensorFlow applications and services.
🔧 Temporary Workarounds
Input Validation
allAdd validation to ensure LookupTableImportV2 is not called with scalar values
# Python code example:
import tensorflow as tf
import numpy as np
# Before calling LookupTableImportV2, check inputs:
def safe_lookup_table_import(table_handle, keys, values):
if values.shape.ndims == 0: # scalar check
raise ValueError('Scalar values not supported for LookupTableImportV2')
return tf.raw_ops.LookupTableImportV2(table_handle=table_handle, keys=keys, values=values)
🧯 If You Can't Patch
- Implement input validation to prevent scalar values from reaching LookupTableImportV2
- Monitor for application crashes and implement automatic restart mechanisms
🔍 How to Verify
Check if Vulnerable:
Check TensorFlow version: 'python -c "import tensorflow as tf; print(tf.__version__)"' - if version is <2.12.0 and not 2.11.1, you are vulnerable.Check Version:
python -c "import tensorflow as tf; print('TensorFlow version:', tf.__version__)"Verify Fix Applied:
After update, verify version is 2.12.0 or 2.11.1 and test LookupTableImportV2 with scalar inputs to ensure no NPE occurs.📡 Detection & Monitoring
Log Indicators:
- Null Pointer Exception (NPE) errors in TensorFlow logs
- Application crashes when using LookupTableImportV2
Network Indicators:
- Unusual termination of TensorFlow services
SIEM Query:
source="tensorflow.logs" AND ("NullPointerException" OR "NPE" OR "LookupTableImportV2" AND "crash")🔗 References
- https://github.com/tensorflow/tensorflow/commit/980b22536abcbbe1b4a5642fc940af33d8c19b69
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-94mm-g2mv-8p7r
- https://github.com/tensorflow/tensorflow/commit/980b22536abcbbe1b4a5642fc940af33d8c19b69
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-94mm-g2mv-8p7r
