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 2876 results — no patch
HIGHEXPLOIT AVAIL

Amnezia-Web-Panel: default credentials in backup download

CVE-2026-108783
7.3
EPSS 0.4%
Auth BypassData LeakageFramework
Amnezia-Web-Panel5143 ATLAS
MEDIUM

Zephyr GT911: stack overflow via rogue touch controller

CVE-2026-19576
6.8
EPSS 0.2%
Code ExecutionDoSFrameworkInference
3 ATLAS
MEDIUM

Pinclaw: missing auth enables agent prompt injection

CVE-2026-108757
6.5
EPSS 0.2%
Auth BypassPrompt InjectionPluginAgent
openclawCWE-30634 ATLAS
MEDIUM

JupyterHub: OAuth client ID collision lets users hijack servers

CVE-2026-108752
4.2
EPSS 0.2%
Auth BypassDoSFramework
CWE-6942 ATLAS
MEDIUM

Quarkus LangChain4j: WebSocket memory leak enables DoS

CVE-2026-108748
5.3
EPSS 0.4%
DoSFrameworkAgent
CWE-4012 ATLAS
MEDIUMEXPLOIT AVAIL

LiteLLM: improper authorization in get_secret api_key

CVE-2026-108575
6.3
EPSS 0.2%
Auth BypassData LeakageInferenceAPIFramework
LiteLLMCWE-26663 ATLAS
MEDIUMEXPLOIT AVAIL

LiteLLM: IDOR in spend logs exposes session data

CVE-2026-108574
4.3
EPSS 0.2%
Auth BypassData LeakageAPIInferenceFramework
LiteLLMCWE-28563 ATLAS
MEDIUM

Cohere SDK: tar slip in SageMaker S3 model extraction

CVE-2026-108597
4.8
EPSS 0.3%
Supply ChainCode ExecutionFrameworkAPI
cohereCWE-223803 ATLAS
MEDIUMEXPLOIT AVAIL

pdf2zh: unauthenticated SSRF via Link input

CVE-2026-108554
5.3
EPSS 0.3%
Data LeakageData ExtractionFrameworkAPI
pdf2zh13 ATLAS
MEDIUM

Jexactyl: IDOR in billing endpoint allows renewing others' servers

CVE-2026-107854
5.4
EPSS 0.2%
Auth BypassFramework
CWE-6392 ATLAS
UNKNOWNEXPLOIT AVAIL

Jexactyl: Stripe payment validation bypass allows underpayment

CVE-2026-107852
--
EPSS 0.2%
Auth BypassSocial EngineeringFrameworkAPI
CWE-3453 ATLAS
HIGHEXPLOIT AVAIL

LobsterAI: skill _meta.json path abuse deletes home dir

CVE-2026-108156
7.1
EPSS 0.2%
Supply ChainSocial EngineeringAgentPlugin
CWE-733 ATLAS
HIGH

Apache CXF: XML signature wrapping bypass in JAX-RS

CVE-2026-100227
7.5
EPSS 0.3%
Auth BypassSupply ChainAPIFramework
1 ATLAS
UNKNOWN

Docker Sandboxes: fail-open proxy leaks OAuth and API tokens

CVE-2026-101998
--
EPSS 0.2%
Data LeakageData ExtractionAgentAPIFramework
anthropic6.3K4 ATLAS
UNKNOWN

Ollama: path traversal in /api/pull enables root RCE

CVE-2026-103663
--
EPSS 0.7%
Code ExecutionSupply ChainInferenceAPIFramework
ollama1.9K2 ATLAS
HIGH

Cyrus SASL: DIGEST-MD5 heap overflow crashes RHOAI clients

CVE-2026-107161
7.5
DoSSupply ChainFrameworkInference
rhoai/odh-mlflow-operator-rhel96952 ATLAS
HIGH

Claude Code: TOCTOU symlink race allows out-of-sandbox write

CVE-2026-103435
--
EPSS 0.3%
Code ExecutionAuth BypassAgentFramework
@anthropic-ai/claude-code3 ATLAS
HIGH

Langflow: access control flaw in vertex cache leaks data

CVE-2026-97680
8.3
EPSS 0.3%
Data LeakageAuth BypassData ExtractionFrameworkRAGAgent
langflowCWE-2844 ATLAS
HIGH

Langflow: authenticated code injection allows RCE

CVE-2026-97679
8.8
EPSS 0.7%
Code ExecutionFrameworkAgent
langflowCWE-943 ATLAS
HIGH

Langflow: authenticated RCE via improper input validation

CVE-2026-97678
8.8
EPSS 0.6%
Code ExecutionFrameworkAgent
langflowCWE-6933 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