CVE-2022-41885: TensorFlow FusedResizeAndPadConv2D

CISO Take

An unauthenticated remote attacker can crash any TensorFlow inference service by sending a request with an oversized tensor shape to FusedResizeAndPadConv2D, causing an integer overflow and service termination. Any internet-exposed TF Serving endpoint running image-based models is affected. Patch immediately to TensorFlow 2.11, 2.10.1, 2.9.3, or 2.8.4 and enforce tensor shape limits at your API gateway.

What is the risk?

Effective risk is HIGH for organizations running public-facing TensorFlow Serving endpoints. CVSS 7.5 reflects network-reachable, zero-auth, low-complexity exploitation with full availability impact. The attack requires no ML knowledge—just a crafted HTTP request with large tensor dimensions. Mitigating factors: no code execution or data exfiltration possible; impact is limited to availability. Not in CISA KEV and no public exploit observed as of disclosure, but PoC is implicit in the advisory.

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

0.4%

chance of exploitation in 30 days

Higher than 34% 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?

AV Network

AC Low

PR None

UI None

S Unchanged

C None

I None

A High

What should I do?

5 steps

PATCH

Upgrade TensorFlow to 2.11.0, or cherry-pick to 2.10.1 / 2.9.3 / 2.8.4. Commit: d66e1d568275e6a2947de97dca7a102a211e01ce.
INPUT VALIDATION

Enforce max tensor shape bounds at API ingress before passing to TF ops—reject requests with dimensions exceeding operational maximums.
API GATEWAY

Deploy rate limiting and request schema validation in front of TF Serving endpoints.
DETECTION

Monitor for abnormal OOM crashes or rapid process restarts in TF Serving logs as an indicator of exploitation attempts.
ISOLATION

Run TF Serving in containers with restart policies and resource limits to minimize availability impact.

What does CISA's SSVC say?

Decision Track*

Exploitation poc

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.T0010.001 - AI Software 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.9 - Risk management system for high-risk AI

ISO 42001

A.9.8 - AI system vulnerability management

NIST AI RMF

GOVERN-1.7 - Processes for AI risk identification and tracking MANAGE-2.4 - Residual risks and recovery procedures

Frequently Asked Questions

What is CVE-2022-41885?

An unauthenticated remote attacker can crash any TensorFlow inference service by sending a request with an oversized tensor shape to FusedResizeAndPadConv2D, causing an integer overflow and service termination. Any internet-exposed TF Serving endpoint running image-based models is affected. Patch immediately to TensorFlow 2.11, 2.10.1, 2.9.3, or 2.8.4 and enforce tensor shape limits at your API gateway.

Is CVE-2022-41885 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2022-41885, increasing the risk of exploitation.

How to fix CVE-2022-41885?

1. PATCH: Upgrade TensorFlow to 2.11.0, or cherry-pick to 2.10.1 / 2.9.3 / 2.8.4. Commit: d66e1d568275e6a2947de97dca7a102a211e01ce. 2. INPUT VALIDATION: Enforce max tensor shape bounds at API ingress before passing to TF ops—reject requests with dimensions exceeding operational maximums. 3. API GATEWAY: Deploy rate limiting and request schema validation in front of TF Serving endpoints. 4. DETECTION: Monitor for abnormal OOM crashes or rapid process restarts in TF Serving logs as an indicator of exploitation attempts. 5. ISOLATION: Run TF Serving in containers with restart policies and resource limits to minimize availability impact.

What systems are affected by CVE-2022-41885?

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

What is the CVSS score for CVE-2022-41885?

CVE-2022-41885 has a CVSS v3.1 base score of 7.5 (HIGH). The EPSS exploitation probability is 0.43%.

What is the AI security impact?

Affected AI Architectures

model servingimage processing pipelinestraining pipelines

MITRE ATLAS Techniques

AML.T0010.001 AI Software

AML.T0029 Denial of AI Service

AML.T0049 Exploit Public-Facing Application

Compliance Controls Affected

EU AI Act: Art.9

ISO 42001: A.9.8

NIST AI RMF: GOVERN-1.7, MANAGE-2.4

What are the technical details?

Original Advisory

TensorFlow is an open source platform for machine learning. When `tf.raw_ops.FusedResizeAndPadConv2D` is given a large tensor shape, it overflows. We have patched the issue in GitHub commit d66e1d568275e6a2947de97dca7a102a211e01ce. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.

Exploitation Scenario

An adversary identifies a public-facing computer vision API (e.g., an image classification service built on TensorFlow Serving). They craft a gRPC or REST request invoking a model that uses FusedResizeAndPadConv2D with tensor shape dimensions set to values near INT32_MAX. The integer overflow triggers a crash in the serving process, causing a denial of service. If the service lacks restart supervision or resource isolation, this can result in prolonged outage. The attack requires no authentication, no ML expertise, and can be repeated to sustain the DoS.

Weaknesses (CWE)

CWE-131 Incorrect Calculation of Buffer Size Primary CWE-131 Incorrect Calculation of Buffer Size

CWE-131 — Incorrect Calculation of Buffer Size: The product does not correctly calculate the size to be used when allocating a buffer, which could lead to a buffer overflow.

[Implementation] When allocating a buffer for the purpose of transforming, converting, or encoding an input, allocate enough memory to handle the largest possible encoding. For example, in a routine that converts "&" characters to "&" for HTML entity encoding, the output buffer needs to be at least 5 times as large as the input buffer.
[Implementation] Understand the programming language's underlying representation and how it interacts with numeric calculation (CWE-681). Pay close attention to byte size discrepancies, precision, signed/unsigned distinctions, truncation, conversion and casting between types, "not-a-number" calculations, and how the language handles numbers that are too large or too small for its underlying representation. [REF-7] Also be careful to account for 32-bit, 64-bit, and other potential differences that may affect the numeric representation.

Source: MITRE CWE corpus.