AVID-2026-R0832

Description

Division by 0 in ResourceScatterDiv in TensorFlow (CVE-2021-37642)

Details

TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation of tf.raw_ops.ResourceScatterDiv is vulnerable to a division by 0 error. The implementation uses a common class for all binary operations but fails to treat the division by 0 case separately. We have patched the issue in GitHub commit 4aacb30888638da75023e6601149415b39763d76. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.

Reason for inclusion in AVID: CVE-2021-37642 describes a division-by-zero vulnerability in TensorFlow’s ResourceScatterDiv kernel. TensorFlow is a core AI framework, making this a vulnerability in software used to build/train/deploy AI systems. It affects AI software stacks and is a documented CVE with a fix, indicating a legitimate security concern within the AI supply chain. The report provides explicit details (affected versions, fix commit, CVSS metrics), supporting its classification as a security and AI-related supply chain issue.

References

• NVD entry
• https://github.com/tensorflow/tensorflow/security/advisories/GHSA-ch4f-829c-v5pw
• https://github.com/tensorflow/tensorflow/commit/4aacb30888638da75023e6601149415b39763d76

Affected or Relevant Artifacts

• Developer: tensorflow
• Deployer: tensorflow
• Artifact Details:

Impact

AVID Taxonomy Categorization

• Risk domains: Security
• SEP subcategories: S0100: Software Vulnerability
• Lifecycle stages: L06: Deployment

CVSS

CWE

Other information

• Report Type: Advisory
• Credits:
• Date Reported: 2021-08-12
• Version: 0.3.3
• AVID Entry