AI Threat Alert

AI Security Threat Feed

Latest CVEs affecting AI/ML systems — LLM frameworks, ML libraries, AI agents, vector databases, and inference servers. Vulnerabilities are tracked from NVD, GitHub Advisory, CISA KEV, MITRE ATLAS, and enriched with CVSS, EPSS, exploitation confidence, AI-component classification, and compliance mappings to ISO 42001, EU AI Act, NIST AI RMF, and OWASP LLM Top 10. Updated continuously as new CVEs are published.

Sources
Each CVE is enriched with
  • CVSS severity
  • EPSS exploit probability
  • Exploitation confidence
  • AI-component classification
  • Compliance mappings
4,307

AI/ML CVEs Tracked

533

Critical

96

New This Week

25

In CISA KEV

Sort:DateCVSSEPSSKEV firstMost exploited
Filter:AllCriticalHighMediumKEV onlyActive exploitationHas patchNo patch

Latest AI Security Threats

Showing 20 of 544 results — Medium severity, has patch
MEDIUM

vllm: ReDoS in inference endpoints enables DoS

GHSA-j828-28rj-hfhp
4.3
DoSInferenceFramework
vllmPatch: 0.9.0CWE-1333943 ATLAS
MEDIUM

vLLM: DoS via unbounded XGrammar schema cache

GHSA-hf3c-wxg2-49q9
6.5
DoSSupply ChainInferenceFrameworkAPI
vllmPatch: 0.8.4CWE-770945 ATLAS
MEDIUM

xgrammar: unbounded grammar cache causes LLM server DoS

CVE-2025-32381
6.5
EPSS 0.5%
DoSInferenceFramework
xgrammarPatch: 0.1.18CWE-7701323 ATLAS
MEDIUM

picklescan: bypass allows silent RCE in ML pipelines

GHSA-v7x6-rv5q-mhwc
--
Supply ChainCode ExecutionModelFramework
picklescanPatch: 0.0.25CWE-18436 ATLAS
MEDIUM

picklescan: numpy bypass enables RCE in ML model pipelines

GHSA-fj43-3qmq-673f
--
Supply ChainCode ExecutionFrameworkModel
picklescanPatch: 0.0.25CWE-50237 ATLAS
MEDIUMEXPLOIT AVAIL

PyTorch: lstm_cell memory corruption, local code exec

CVE-2025-3001
5.3
EPSS 0.2%
Code ExecutionDoSFramework
torchPatch: 2.10.0CWE-11926.1K3 ATLAS
MEDIUMEXPLOIT AVAIL

PyTorch: memory corruption in torch.jit.script compiler

CVE-2025-3000
5.3
EPSS 0.2%
Code ExecutionSupply ChainFrameworkInference
torchPatch: 2.13.0CWE-11926.1K3 ATLAS
MEDIUMEXPLOIT AVAIL

PyTorch: memory corruption in RNN sequence unpacking

CVE-2025-2999
5.3
EPSS 0.2%
Code ExecutionData ExtractionSupply ChainFrameworkTraining DataInference
torchPatch: 2.9.1CWE-11926.1K3 ATLAS
MEDIUMEXPLOIT AVAIL

SageMaker SDK: MD5 collision silently replaces ML workflows

CVE-2025-0508
5.9
EPSS 0.3%
Supply ChainModel PoisoningFrameworkTraining Data
sagemakerPatch: 2.237.3CWE-328503 ATLAS
MEDIUMEXPLOIT AVAIL

llama-index: DoS via infinite recursion in web reader

CVE-2024-12910
5.9
EPSS 0.7%
DoSSupply ChainFrameworkRAG
llama-indexPatch: 0.12.9CWE-4002514 ATLAS
MEDIUMEXPLOIT AVAIL

langchain-core: file read via prompt template inputs

CVE-2024-10940
5.3
EPSS 0.4%
Data ExtractionData LeakagePrivacy ViolationFrameworkAgent
langchain-corePatch: 0.1.53CWE-4976.3K4 ATLAS
MEDIUMEXPLOIT AVAIL

picklescan: ZIP spoof lets malicious PyTorch models bypass scan

CVE-2025-1944
6.5
EPSS 0.3%
Supply ChainCode ExecutionFrameworkModel
picklescanPatch: 0.0.23CWE-34535 ATLAS
MEDIUM

Ray: Redis password exposed via plaintext logging

CVE-2025-1979
6.4
EPSS 0.2%
Data LeakageData ExtractionFrameworkTraining Data
rayPatch: 2.43.0CWE-5326424 ATLAS
MEDIUMEXPLOIT AVAIL

picklescan: scanner bypass enables supply chain RCE

CVE-2025-1716
--
EPSS 1.7%
Supply ChainCode ExecutionModelFramework
picklescanPatch: 0.0.22CWE-18435 ATLAS
MEDIUMEXPLOIT AVAIL

picklescan: extension bypass enables RCE on model load

CVE-2025-1889
--
EPSS 0.4%
Supply ChainCode ExecutionModelFramework
picklescanPatch: 0.0.22CWE-64635 ATLAS
MEDIUMEXPLOIT AVAILSCANNER

Label Studio: reflected XSS via label_config param

CVE-2025-25296
6.1
EPSS 1.9%
Code ExecutionData ExtractionTraining DataFramework
label-studioPatch: 1.16.0CWE-7924 ATLAS
MEDIUMEXPLOIT AVAIL

Composio: command injection in AI agent tool calls

CVE-2024-53526
6.4
EPSS 0.6%
Code ExecutionSupply ChainFrameworkAgentAPI
composio-openaiPatch: 0.6.9CWE-7715.6K5 ATLAS
MEDIUM

Giskard: ReDoS in text perturbation causes DoS

CVE-2024-52524
--
EPSS 0.8%
DoSSupply ChainFramework
giskardPatch: 2.15.5CWE-13333 ATLAS
MEDIUM

Gradio: Dropdown validation bypass enables arbitrary input

GHSA-26jh-r8g2-6fpr
5.3
Auth BypassCode ExecutionFrameworkInferenceAPI
gradioPatch: 5.0.07353 ATLAS
MEDIUMEXPLOIT AVAIL

open-webui: IDOR enables cross-user memory tampering

CVE-2024-7041
6.5
EPSS 0.4%
Auth BypassModel PoisoningPrivacy ViolationAPIAgentFramework
open-webuiPatch: 0.6.3CWE-25024 ATLAS

Frequently asked questions

What is an AI security threat feed?

An AI security threat feed is a continuously updated stream of vulnerabilities (CVEs) affecting AI and machine-learning systems — LLM frameworks, ML libraries, AI agents, vector databases, and inference servers — filtered out of the broader CVE firehose and enriched for relevance.

Which sources are the AI CVEs tracked from?

CVEs are tracked from NVD, GitHub Advisory, CISA KEV, and MITRE ATLAS, then enriched with CVSS, EPSS, exploitation confidence, AI-component classification, and compliance mappings.

What AI systems do these vulnerabilities affect?

Coverage spans LLM frameworks, ML libraries, AI agents, vector databases, and inference servers — the components most security teams now run in production.

How often is the AI threat feed updated?

The feed updates continuously as new CVEs are published and enriched, so the most recent AI/ML vulnerabilities appear at the top.

Is the AI security feed free?

Yes — the public feed is free to browse. A Pro subscription adds breaking alerts, MITRE ATLAS mappings, compliance reports (ISO 42001, EU AI Act), and full CISO analysis.

Need deeper analysis?

Get ATLAS technique mappings, compliance reports (ISO 42001, EU AI Act), breaking alerts, and full CISO analysis with a Pro subscription.

Start 14-Day Free Trial