CVE-2022-29198: TensorFlow: DoS via sparse tensor input validation failure

MEDIUMPoC AVAILABLECISA: TRACK*

Published May 20, 2022

CISO Take

This medium-severity DoS in TensorFlow crashes the runtime via malformed sparse tensor inputs—no RCE, no data leakage. Risk is real in shared ML environments (Jupyter clusters, multi-tenant GPU nodes) where low-privileged users can submit crafted inputs. Patch to TF 2.9.0/2.8.1/2.7.2/2.6.4 at next maintenance window; no emergency action required.

What is the risk?

CVSS 5.5 Medium with local attack vector and low privilege requirement limits blast radius. Exploitation requires access to the TF runtime, which in practice means shared notebooks, Kubeflow pipelines, or on-prem GPU clusters with multi-user access. No evidence of active exploitation and not in CISA KEV. Severity elevates in high-availability inference environments where process crashes affect SLAs, but remains low-priority compared to RCE or data exposure vectors.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
TensorFlow	pip	—	No patch
201.3KOpenSSF 7.63.4K dependentsPushed 5d ago4% patched~1372d to patchFull package profile →

Do you use TensorFlow? You're affected.

How severe is it?

CVSS 3.1

5.5 / 10

EPSS

0.3%

chance of exploitation in 30 days

Higher than 26% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Trivial

Exploitation Confidence

medium

○CISA SSVC: Public PoC

○Public PoC indexed (trickest/cve)

Composite signal derived from CISA KEV, VulnCheck KEV, CISA SSVC, EPSS, Metasploit, Exploit-DB, trickest/cve, Nuclei templates, and inthewild.io exploitation reports.

What is the attack surface?

AVLocal

ACLow

PRLow

UINone

SUnchanged

CNone

INone

AHigh

What should I do?

5 steps

Upgrade TensorFlow to 2.9.0, 2.8.1, 2.7.2, or 2.6.4.
If immediate patching is not possible, validate that dense_shape is rank-1 and indices is rank-2 before calling SparseTensorToCSRSparseMatrix at the application layer.
In shared environments, restrict which users can submit raw ops to the TF runtime.
For inference services, add input shape validation at the API gateway before tensors reach TF.
Monitor for unexpected TF process crashes—repeated CHECK-failure crashes with user-controlled inputs indicate active exploitation attempts.

What does CISA's SSVC say?

DecisionTrack*

Exploitationpoc

AutomatableNo

Technical Impactpartial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

DoSFrameworkInferenceAML.T0029- Denial of AI Service AML.T0034- Cost Harvesting AML.T0049- Exploit Public-Facing Application

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Article 15-Accuracy, Robustness and Cybersecurity

ISO 42001

A.6.1.4-AI System Availability and Resilience

NIST AI RMF

MANAGE-2.2-Risk Response: Availability and Resilience

OWASP LLM Top 10

LLM04-Model Denial of Service

Frequently Asked Questions

What is CVE-2022-29198?

Is CVE-2022-29198 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2022-29198, increasing the risk of exploitation.

How to fix CVE-2022-29198?

1. Upgrade TensorFlow to 2.9.0, 2.8.1, 2.7.2, or 2.6.4. 2. If immediate patching is not possible, validate that dense_shape is rank-1 and indices is rank-2 before calling SparseTensorToCSRSparseMatrix at the application layer. 3. In shared environments, restrict which users can submit raw ops to the TF runtime. 4. For inference services, add input shape validation at the API gateway before tensors reach TF. 5. Monitor for unexpected TF process crashes—repeated CHECK-failure crashes with user-controlled inputs indicate active exploitation attempts.

What systems are affected by CVE-2022-29198?

This vulnerability affects the following AI/ML architecture patterns: training pipelines, model serving, inference, shared ML platforms.

What is the CVSS score for CVE-2022-29198?

CVE-2022-29198 has a CVSS v3.1 base score of 5.5 (MEDIUM). The EPSS exploitation probability is 0.33%.

What is the AI security impact?

Affected AI Architectures

training pipelinesmodel servinginferenceshared ML platforms

MITRE ATLAS Techniques

AML.T0029Denial of AI Service

AML.T0034Cost Harvesting

AML.T0049Exploit Public-Facing Application

Compliance Controls Affected

EU AI Act: Article 15

ISO 42001: A.6.1.4

NIST AI RMF: MANAGE-2.2

OWASP LLM Top 10: LLM04

What are the technical details?

Original Advisory

TensorFlow is an open source platform for machine learning. Prior to versions 2.9.0, 2.8.1, 2.7.2, and 2.6.4, the implementation of `tf.raw_ops.SparseTensorToCSRSparseMatrix` does not fully validate the input arguments. This results in a `CHECK`-failure which can be used to trigger a denial of service attack. The code assumes `dense_shape` is a vector and `indices` is a matrix (as part of requirements for sparse tensors) but there is no validation for this. Versions 2.9.0, 2.8.1, 2.7.2, and 2.6.4 contain a patch for this issue.

Exploitation Scenario

An attacker with access to a shared Jupyter Hub instance or Kubeflow pipeline crafts a SparseTensor where dense_shape is a matrix instead of a vector, or indices is a vector instead of a matrix. Passing this to tf.raw_ops.SparseTensorToCSRSparseMatrix triggers an unchecked assertion failure, crashing the TF process. In a multi-tenant ML platform, this disrupts other users' training jobs sharing the same runtime. In an inference API exposing sparse tensor endpoints, repeated calls constitute a targeted availability attack without rate limiting.

Weaknesses (CWE)

CWE-20Improper Input Validation

CWE-20 — Improper Input Validation: The product receives input or data, but it does not validate or incorrectly validates that the input has the properties that are required to process the data safely and correctly.

[Architecture and Design] Consider using language-theoretic security (LangSec) techniques that characterize inputs using a formal language and build "recognizers" for that language. This effectively requires parsing to be a distinct layer that effectively enforces a boundary between raw input and internal data representations, instead of allowing parser code to be scattered throughout the program, where it could be subject to errors or inconsistencies that create weaknesses. [REF-1109] [REF-1110] [REF-1111]
[Architecture and Design] Use an input validation framework such as Struts or the OWASP ESAPI Validation API. Note that using a framework does not automatically address all input validation problems; be mindful of weaknesses that could arise from misusing the framework itself (CWE-1173).

Source: MITRE CWE corpus.