CVE-2022-23562 — HIGH (CVSS 8.8) AI Security Vulnerability

CISO Take

Upgrade TensorFlow immediately to 2.8.0, 2.7.1, 2.6.3, or 2.5.3. Any deployment accepting user-controlled tensor data over a network is exploitable by low-privileged attackers with no user interaction required. Prioritize inference APIs and multi-tenant ML platforms where untrusted inputs reach the framework.

Risk Assessment

High risk (CVSS 8.8). Network-accessible, low complexity, low privileges required — exploitable remotely with minimal skill. The integer overflow in the Range op triggers undefined behavior (potential RCE) or massive memory allocations (DoS). TensorFlow's ubiquity in production ML infrastructure significantly widens the exposure surface. No active exploitation reported and not in CISA KEV, but the low barrier to exploitation demands prompt patching.

Affected Systems

Package	Ecosystem	Vulnerable Range	Patched
tensorflow	pip	—	No patch
195.0K OpenSSF 7.2 3.7K dependents Pushed 6d ago 4% patched ~1372d to patch Full package profile →

Do you use tensorflow? You're affected.

Severity & Risk

CVSS 3.1

8.8 / 10

EPSS

0.4%

chance of exploitation in 30 days

Higher than 58% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

No known exploitation

Sophistication

Moderate

Attack Surface

AV Network

AC Low

PR Low

UI None

S Unchanged

C High

I High

A High

Recommended Action

4 steps

Patch: Upgrade to TensorFlow 2.8.0 or apply cherry-picks in 2.7.1/2.6.3/2.5.3.
Workaround: Enforce bounds validation on all integer inputs to tf.range() at API boundaries — reject start/limit/delta values outside sane operational ranges before they reach TensorFlow.
Detection: Alert on anomalous memory spikes and OOM errors in TensorFlow processes; audit inference API logs for malformed tensor operation requests.
Isolation: Run TensorFlow serving containers with memory limits (--memory flag) to cap DoS blast radius and restrict process permissions to limit RCE impact.

Classification

Code Execution DoS Framework Inference AML.T0010.001 - AI Software AML.T0029 - Denial of AI Service AML.T0049 - Exploit Public-Facing Application

Compliance Impact

This CVE is relevant to:

EU AI Act

Art. 15 - Accuracy, Robustness and Cybersecurity

ISO 42001

A.6.2 - AI system risk assessment

NIST AI RMF

MANAGE-2.2 - AI Risk Response

OWASP LLM Top 10

LLM05:2025 - Supply Chain Vulnerabilities

Frequently Asked Questions

What is CVE-2022-23562?

Upgrade TensorFlow immediately to 2.8.0, 2.7.1, 2.6.3, or 2.5.3. Any deployment accepting user-controlled tensor data over a network is exploitable by low-privileged attackers with no user interaction required. Prioritize inference APIs and multi-tenant ML platforms where untrusted inputs reach the framework.

Is CVE-2022-23562 actively exploited?

No confirmed active exploitation of CVE-2022-23562 has been reported, but organizations should still patch proactively.

How to fix CVE-2022-23562?

1. Patch: Upgrade to TensorFlow 2.8.0 or apply cherry-picks in 2.7.1/2.6.3/2.5.3. 2. Workaround: Enforce bounds validation on all integer inputs to tf.range() at API boundaries — reject start/limit/delta values outside sane operational ranges before they reach TensorFlow. 3. Detection: Alert on anomalous memory spikes and OOM errors in TensorFlow processes; audit inference API logs for malformed tensor operation requests. 4. Isolation: Run TensorFlow serving containers with memory limits (--memory flag) to cap DoS blast radius and restrict process permissions to limit RCE impact.

What systems are affected by CVE-2022-23562?

This vulnerability affects the following AI/ML architecture patterns: model serving, training pipelines, inference API, ML platforms, notebook/Jupyter environments.

What is the CVSS score for CVE-2022-23562?

CVE-2022-23562 has a CVSS v3.1 base score of 8.8 (HIGH). The EPSS exploitation probability is 0.36%.

Technical Details

NVD Description

Tensorflow is an Open Source Machine Learning Framework. The implementation of `Range` suffers from integer overflows. These can trigger undefined behavior or, in some scenarios, extremely large allocations. The fix will be included in TensorFlow 2.8.0. We will also cherrypick this commit on TensorFlow 2.7.1, TensorFlow 2.6.3, and TensorFlow 2.5.3, as these are also affected and still in supported range.

Exploitation Scenario

An adversary with a low-privileged account on a TensorFlow-based model serving platform submits a crafted inference request containing a Range operation with carefully chosen integer values designed to overflow. This triggers either a massive heap allocation that crashes the inference server (taking down all users) or memory corruption that, under exploitable memory layout conditions, enables code execution within the TF serving process — potentially exfiltrating loaded model weights, API keys in environment variables, or establishing a foothold in the ML infrastructure.