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: chat template bypass enables DoS via /v1/responses

CVE-2026-115918
6.5
EPSS 0.4%
DoSInferenceAPI
vllmPatch: 0.32.0CWE-400943 ATLAS
MEDIUM

vLLM: off-by-one token ID check crashes engine (DoS)

CVE-2026-115916
6.5
EPSS 0.3%
DoSInferenceFramework
vllmPatch: 0.32.0CWE-129942 ATLAS
MEDIUM

vLLM: Phi-4 audio attention causes GPU OOM DoS

CVE-2026-115913
6.5
EPSS 0.3%
DoSInferenceModel
vllmPatch: 0.32.0CWE-770943 ATLAS
MEDIUM

vLLM: chat template DoS via unbounded Jinja rendering

CVE-2026-115911
6.5
EPSS 0.3%
DoSSupply ChainInferenceAPIFramework
vllmPatch: 0.32.0CWE-400944 ATLAS
MEDIUM

vLLM: zero-sum OCR image crashes engine (DoS)

CVE-2026-115914
6.5
EPSS 0.4%
DoSInferenceAPI
vllmPatch: 0.32.0CWE-617943 ATLAS
MEDIUM

filelock: rename-to-rmtree deletes other users' dirs

CVE-2026-112937
6.1
EPSS 0.2%
DoSSupply ChainFramework
filelockPatch: 4.2.1CWE-3676.3K1 ATLAS
MEDIUMEXPLOIT AVAIL

virtualenv: stale temp-cache symlink enables local code exec

CVE-2026-113713
--
EPSS 0.2%
Code ExecutionSupply ChainFramework
virtualenvPatch: 21.14.10CWE-3776.3K2 ATLAS
MEDIUM

virtualenv: cached file modes expose seeded envs to tampering

CVE-2026-113720
--
EPSS 0.1%
Code ExecutionSupply ChainFramework
virtualenvPatch: 21.14.9CWE-7326.3K1 ATLAS
MEDIUMEXPLOIT AVAIL

virtualenv: path traversal in seed cleanup deletes files

CVE-2026-113719
4.5
EPSS 0.2%
Supply ChainFramework
virtualenvPatch: 21.14.9CWE-226.3K2 ATLAS
MEDIUMEXPLOIT AVAILSCANNER

MLflow: unsafe scorer deserialization enables SSRF

CVE-2026-115886
6.5
EPSS 0.7%
Auth BypassData LeakageFrameworkAPIInference
mlflowPatch: 3.17.0CWE-4706953 ATLAS
MEDIUMEXPLOIT AVAIL

filelock: soft-lock inode-reuse race allows double lock

CVE-2026-112906
4.7
EPSS 0.2%
Supply ChainFramework
filelockPatch: 4.2.1CWE-6946.3K2 ATLAS
MEDIUM

filelock: symlink race in stale-lock restore links files

CVE-2026-112914
5.3
EPSS 0.2%
Data LeakageSupply ChainFramework
filelockPatch: 4.2.1CWE-596.3K2 ATLAS
MEDIUM

FastMCP: post-transform authorization skip exposes tools

GHSA-7xgj-8392-5q4r
6.8
Auth BypassData LeakageFrameworkAgentPlugin
fastmcp-slimPatch: 3.4.10CWE-8621.8K3 ATLAS
MEDIUMEXPLOIT AVAIL

SiYuan: auth oracle leaks hidden document existence

CVE-2026-73630
5.8
EPSS 0.3%
Data LeakagePrivacy ViolationFrameworkAPI
github.com/siyuan-note/siyuan/kernelPatch: 0.0.0-20260731093152-c7522b8db586CWE-2033.0K2 ATLAS
MEDIUM

SiYuan: hidden-tier docs leak via AV backlinks API

CVE-2026-73049
5.8
EPSS 0.3%
Data LeakageAuth BypassFrameworkAPI
github.com/siyuan-note/siyuan/kernelPatch: 0.0.0-20260730150429-5b42614311b0CWE-8635142 ATLAS
MEDIUM

SiYuan: publish storage endpoint leaks admin context

GHSA-xp3q-r38w-vgqm
5.8
Data LeakageAuth BypassPrivacy ViolationFrameworkAPI
github.com/siyuan-note/siyuan/kernelPatch: 0.0.0-20260729005258-4daee87d4152CWE-8633.0K3 ATLAS
MEDIUM

virtualenv: symlink following allows file overwrite

CVE-2026-113721
4.5
EPSS 0.1%
Supply ChainCode ExecutionFramework
virtualenvPatch: 21.14.9CWE-596.3K2 ATLAS
MEDIUM

virtualenv: symlink following allows writes outside venv

CVE-2026-113716
4.5
EPSS 0.1%
Supply ChainCode ExecutionFramework
virtualenvPatch: 21.14.9CWE-596.3K2 ATLAS
MEDIUM

virtualenv: symlink-following chmod alters file permissions

CVE-2026-113724
4.5
EPSS 0.2%
Data LeakagePrivacy ViolationFramework
virtualenvPatch: 21.14.9CWE-596.3K2 ATLAS
MEDIUM

Pydantic AI Harness: tool-call limit bypass in CodeMode

CVE-2026-114386
4.8
EPSS 0.3%
DoSPrompt InjectionFrameworkAgentPlugin
pydantic-ai-harnessPatch: 0.59.0CWE-7705124 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