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
2,375

AI/ML CVEs Tracked

336

Critical

408

New This Week

18

In CISA KEV

Sort: Date CVSS EPSS KEV first Most exploited
Filter: All Critical High Medium KEV only Active exploitation Has patch No patch

Latest AI Security Threats

Showing 20 of 1553 results — no patch
HIGH EXPLOIT AVAIL

PickleScan: pty.spawn bypass enables RCE in model scans

CVE-2025-71322
8.8
EPSS 0.4%
Supply Chain Code Execution Model Framework
PickleScan CWE-693 3 5 ATLAS
CRITICAL EXPLOIT AVAIL

picklescan: blocklist bypass allows arbitrary file write/RCE

CVE-2025-71321
9.8
EPSS 0.6%
Supply Chain Code Execution DoS Framework Model Training Data
picklescan 3 4 ATLAS
CRITICAL EXPLOIT AVAIL

picklescan: deny-list bypass enables arbitrary RCE

CVE-2025-71320
9.8
EPSS 0.6%
Supply Chain Code Execution Model Framework
picklescan 3 4 ATLAS
MEDIUM

vLLM: decompression bomb OOM via audio endpoint

CVE-2026-54233
6.5
EPSS 0.0%
DoS Inference API
vllm CWE-409 130 3 ATLAS
MEDIUM

vLLM: heap address leak enables ASLR bypass

CVE-2026-54236
5.3
EPSS 0.0%
Data Leakage Code Execution Inference API
vllm CWE-532 130 3 ATLAS
AWAITING NVD

vLLM: integer truncation leaks GPU memory cross-tenant

CVE-2026-53923
--
EPSS 0.0%
Data Extraction Supply Chain Privacy Violation Inference Model
vllm CWE-200 130 4 ATLAS
MEDIUM

vllm: EXIF/tRNS preprocessing gap enables adversarial input

GHSA-8jr5-v98p-w75m
4.8
Adversarial Examples Prompt Injection Inference Framework
vllm CWE-436 130 4 ATLAS
AWAITING NVD

vLLM: NaN/Inf bypass crashes GPU inference workers

CVE-2026-54235
--
EPSS 0.0%
DoS Inference
vllm CWE-1287 130 3 ATLAS
CRITICAL

Oracle Coherence: unauthenticated HTTP full takeover

CVE-2026-35310
9.8
EPSS 0.5%
Auth Bypass Code Execution Data Extraction API Inference Framework
coherence 653 4 ATLAS
CRITICAL

Oracle Coherence: unauthenticated RCE via HTTP (CVSS 9.8)

CVE-2026-35309
9.8
EPSS 0.5%
Code Execution Auth Bypass Data Extraction Framework API Inference
coherence 653 3 ATLAS
CRITICAL

Oracle Coherence: unauthenticated RCE via third-party jars

CVE-2026-35308
10.0
EPSS 0.5%
Code Execution Auth Bypass Supply Chain Framework Inference API
coherence 653 4 ATLAS
CRITICAL

Oracle Coherence: unauthenticated RCE, CVSS 10.0

CVE-2026-35307
10.0
EPSS 0.5%
Code Execution Auth Bypass Data Extraction Framework API Inference
coherence 653 4 ATLAS
CRITICAL

Oracle Coherence: unauthenticated data access via HTTP

CVE-2026-35306
9.3
EPSS 0.4%
Data Extraction Auth Bypass Data Leakage API Inference Framework
coherence 653 4 ATLAS
CRITICAL

Oracle Coherence: unauth data exfiltration via bundled libs

CVE-2026-35305
9.3
EPSS 0.3%
Data Extraction Auth Bypass API Framework
coherence 653 4 ATLAS
CRITICAL

Oracle Coherence: unauthenticated HTTPS takeover

CVE-2026-35304
9.8
EPSS 0.5%
Auth Bypass Code Execution Data Extraction API Inference Framework
coherence 653 4 ATLAS
MEDIUM

vLLM: image metadata mishandling corrupts multimodal inputs

CVE-2026-12491
4.8
EPSS 0.2%
Adversarial Examples DoS Inference Model
rhaiis/vllm-cpu-rhel9 130 4 ATLAS
HIGH

WPBot Pro: subscriber file deletion → system DoS

CVE-2025-60223
7.7
EPSS 0.4%
DoS Auth Bypass Plugin
wpbot-pro 3 ATLAS
LOW

pi-coding-agent: XSS in HTML exports via prompt injection

CVE-2026-54326
2.5
EPSS 0.0%
Prompt Injection Code Execution Agent
CWE-79 4 ATLAS
CRITICAL

backpropagate: auth bypass exposes LLM training plane

CVE-2026-48797
--
EPSS 0.4%
Auth Bypass Data Extraction DoS Training Data Model Framework
CWE-358 6 ATLAS
MEDIUM

LangGraph SDK: path traversal bypasses proxy-layer authz

CVE-2026-48776
4.2
EPSS 0.2%
Auth Bypass Data Extraction Framework Agent
langchain-ai 2.7K 3 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