CVE-2022-35941: TensorFlow DoS via negative ksize

CISO Take

TensorFlow inference endpoints accepting user-controlled model parameters are vulnerable to remote crash via a negative ksize value in AvgPoolOp. No authentication required—any network-accessible TF Serving instance is at risk. Patch to TensorFlow 2.10.0 (or cherry-picks on 2.9.1/2.8.1/2.7.2) immediately and add positive-integer validation at API boundaries as a defense-in-depth measure.

What is the risk?

High severity (CVSS 7.5) with zero-barrier exploitability—no auth, no user interaction, low complexity over the network. Pure availability impact with no confidentiality or integrity risk. Practical risk concentrates on organizations running TensorFlow Serving or custom inference APIs where request parameters are not sanitized before reaching kernel operations. Widely deployed framework increases blast radius across the ML ecosystem.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
TensorFlow	pip	—	No patch
195.8K OpenSSF 7.1 3.7K dependents Pushed 3d ago 4% patched ~1372d to patch Full package profile →

Do you use TensorFlow? You're affected.

How severe is it?

CVSS 3.1

7.5 / 10

EPSS

0.5%

chance of exploitation in 30 days

Higher than 41% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

No known exploitation

Sophistication

Trivial

What is the attack surface?

AV Network

AC Low

PR None

UI None

S Unchanged

C None

I None

A High

What should I do?

5 steps

PATCH

Upgrade to TensorFlow 2.10.0 or apply cherry-pick commits on 2.9.1, 2.8.1, or 2.7.2. Commit: 3a6ac52664c6c095aa2b114e742b0aa17fdce78f.
VALIDATE

Enforce ksize > 0 at API entry points before parameters reach TensorFlow kernel execution.
ISOLATE

Run TF Serving in containers with auto-restart policies (Kubernetes restartPolicy: Always) to minimize crash-induced downtime.
DETECT

Monitor TF Serving logs and process supervisor metrics for abnormal crash/restart patterns as exploitation indicators.
AUDIT

Inventory all TF deployments exposing AvgPool-based models to external or user-controlled input.

What does CISA's SSVC say?

Decision Track

Exploitation none

Automatable No

Technical Impact partial

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

How is it classified?

DoS Framework Inference AML.T0029 - Denial of AI Service AML.T0049 - Exploit Public-Facing Application

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Art. 15 - Accuracy, robustness and cybersecurity

ISO 42001

A.6.2.5 - AI system availability and resilience

NIST AI RMF

MANAGE-2.2 - Mechanisms to sustain the value of deployed AI systems are evaluated and applied

Frequently Asked Questions

What is CVE-2022-35941?

TensorFlow inference endpoints accepting user-controlled model parameters are vulnerable to remote crash via a negative ksize value in AvgPoolOp. No authentication required—any network-accessible TF Serving instance is at risk. Patch to TensorFlow 2.10.0 (or cherry-picks on 2.9.1/2.8.1/2.7.2) immediately and add positive-integer validation at API boundaries as a defense-in-depth measure.

Is CVE-2022-35941 actively exploited?

No confirmed active exploitation of CVE-2022-35941 has been reported, but organizations should still patch proactively.

How to fix CVE-2022-35941?

1. PATCH: Upgrade to TensorFlow 2.10.0 or apply cherry-pick commits on 2.9.1, 2.8.1, or 2.7.2. Commit: 3a6ac52664c6c095aa2b114e742b0aa17fdce78f. 2. VALIDATE: Enforce ksize > 0 at API entry points before parameters reach TensorFlow kernel execution. 3. ISOLATE: Run TF Serving in containers with auto-restart policies (Kubernetes restartPolicy: Always) to minimize crash-induced downtime. 4. DETECT: Monitor TF Serving logs and process supervisor metrics for abnormal crash/restart patterns as exploitation indicators. 5. AUDIT: Inventory all TF deployments exposing AvgPool-based models to external or user-controlled input.

What systems are affected by CVE-2022-35941?

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

What is the CVSS score for CVE-2022-35941?

CVE-2022-35941 has a CVSS v3.1 base score of 7.5 (HIGH). The EPSS exploitation probability is 0.54%.

What is the AI security impact?

Affected AI Architectures

model servingtraining pipelinesinference endpoints

MITRE ATLAS Techniques

AML.T0029 Denial of AI Service

AML.T0049 Exploit Public-Facing Application

Compliance Controls Affected

EU AI Act: Art. 15

ISO 42001: A.6.2.5

NIST AI RMF: MANAGE-2.2

What are the technical details?

Original Advisory

TensorFlow is an open source platform for machine learning. The `AvgPoolOp` function takes an argument `ksize` that must be positive but is not checked. A negative `ksize` can trigger a `CHECK` failure and crash the program. We have patched the issue in GitHub commit 3a6ac52664c6c095aa2b114e742b0aa17fdce78f. The fix will be included in TensorFlow 2.10.0. We will also cherrypick this commit on TensorFlow 2.9.1, TensorFlow 2.8.1, and TensorFlow 2.7.2, as these are also affected and still in supported range. There are no known workarounds to this issue.

Exploitation Scenario

An attacker identifies a model serving endpoint powered by TensorFlow (via architecture fingerprinting or public documentation). They craft a REST API request to TF Serving with a negative ksize value (e.g., ksize=-1) targeting a model with an AvgPool layer. The CHECK assertion fires immediately, crashing the process. With no rate limiting or input validation, the attacker repeatedly crashes the service—causing sustained availability disruption to AI-powered applications such as fraud detection, recommendation engines, or computer vision pipelines—with zero credentials required.

Weaknesses (CWE)

CWE-617 Reachable Assertion

CWE-617 — Reachable Assertion: The product contains an assert() or similar statement that can be triggered by an attacker, which leads to an application exit or other behavior that is more severe than necessary.

[Implementation] Make sensitive open/close operation non reachable by directly user-controlled data (e.g. open/close resources)
[Implementation] Perform input validation on user data.

Source: MITRE CWE corpus.