AVID-2026-R0898

Description

Heap OOB read in all tf.raw_ops.QuantizeAndDequantizeV* ops (CVE-2021-41205)

Details

TensorFlow is an open source platform for machine learning. In affected versions the shape inference functions for the QuantizeAndDequantizeV* operations 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-2021-41205 describes a heap out-of-bounds read vulnerability in TensorFlow’s QuantizeAndDequantizeV* ops. TensorFlow is an AI framework widely used in AI/ML pipelines, serving as a dependency in many AI systems. The issue affects software components (TensorFlow) used to build/train/deploy AI systems, representing a software supply-chain vulnerability within AI stacks. It is a security vulnerability (out-of-bounds read) with explicit CVSS metadata and a remediation path. Therefore it meets AVID criteria for a GP AI supply-chain vulnerability.

References

Affected or Relevant Artifacts

Developer: tensorflow
Deployer: tensorflow
Artifact Details:

Type	Name
System	tensorflow

Impact

AVID Taxonomy Categorization

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

CVSS

Version	3.1
Vector String	CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H
Base Score	7.1
Base Severity	🔴 High
Attack Vector	LOCAL
Attack Complexity	🟢 Low
Privileges Required	🟢 Low
User Interaction	NONE
Scope	UNCHANGED
Confidentiality Impact	🔴 High
Integrity Impact	NONE
Availability Impact	🔴 High

CWE

ID	Description
CWE-125	CWE-125: Out-of-bounds Read

Other information

Report Type: Advisory
Credits:
Date Reported: 2021-11-05
Version: 0.3.3
AVID Entry