AVID-2026-R0904

Description

Heap OOB read in tf.ragged.cross (CVE-2021-41212)

Details

TensorFlow is an open source platform for machine learning. In affected versions the shape inference code for tf.ragged.cross can trigger a read outside of bounds of heap allocated array. The fix will be included in TensorFlow 2.7.0. We will also cherrypick this commit on TensorFlow 2.6.1, TensorFlow 2.5.2, and TensorFlow 2.4.4, as these are also affected and still in supported range.

Reason for inclusion in AVID: CVE-2021-41212 describes a heap out-of-bounds read in TensorFlow (tf.ragged.cross), a vulnerability in a software library used to develop and run AI/ML models. It affects TensorFlow versions used in AI pipelines and is a conventional CVE with public advisories and references. It relates to software components (ML framework) critical to general-purpose AI systems, not hardware/firmware. The report provides clear evidence (NVD entry, GitHub advisory, commit) and CVSS data, indicating a genuine security vulnerability in the software stack used for AI—thus relevant to the AI supply chain.

References

• NVD entry
• https://github.com/tensorflow/tensorflow/security/advisories/GHSA-fr77-rrx3-cp7g
• https://github.com/tensorflow/tensorflow/commit/fa6b7782fbb14aa08d767bc799c531f5e1fb3bb8

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: 2021-11-05
• Version: 0.3.3
• AVID Entry