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 169 results — Critical severity, has patch
CRITICAL

MLflow: scorer deserialization allows unauthenticated RCE

CVE-2026-116692
9.8
EPSS 0.7%
Code ExecutionSupply ChainAuth BypassFrameworkAPI
mlflowPatch: 3.17.0CWE-946954 ATLAS
CRITICAL

MONAI: NaN/inf DICOM input causes DoS and corrupt inference

CVE-2026-48705
9.3
DoSAdversarial ExamplesFrameworkInferenceAPI
monaiPatch: 1.6.1CWE-201285 ATLAS
CRITICAL

LiteLLM: forged admin token enables RCE via MCP stdio

CVE-2026-114612
9.9
EPSS 0.3%
Auth BypassCode ExecutionAPIFrameworkPlugin
litellmPatch: 1.100.4CWE-26965 ATLAS
CRITICALEXPLOIT AVAIL

Pillow: heap overflow via lossless JPEG in thumbnail()

CVE-2026-112403
--
EPSS 0.9%
DoSCode ExecutionSupply ChainFrameworkInferenceTraining Data
PillowPatch: 13.0.0CWE-1226.3K3 ATLAS
CRITICAL

Mercury: missing auth allows anonymous kernel RCE

CVE-2026-111058
9.8
EPSS 0.5%
Auth BypassCode ExecutionFramework
mercuryPatch: 3.2.6CWE-8623.0K4 ATLAS
CRITICALEXPLOIT AVAIL

Mercury: regex blocklist bypass allows unauthenticated RCE

CVE-2026-111057
10.0
EPSS 0.7%
Auth BypassCode ExecutionFrameworkPlugin
mercuryPatch: 3.2.4CWE-1853.0K4 ATLAS
CRITICAL

Podman: checkpoint images override sandbox on run

CVE-2026-94603
10.0
EPSS 0.5%
Supply ChainCode ExecutionFrameworkInference
github.com/containers/podmanPatch: 5.8.8CWE-6936.3K4 ATLAS
CRITICALEXPLOIT AVAIL

PraisonAI: code injection in deploy API enables RCE

CVE-2026-62176
9.1
EPSS 0.5%
Code ExecutionSupply ChainFrameworkAgentAPI
PraisonAIPatch: 4.6.78CWE-9414 ATLAS
CRITICALEXPLOIT AVAIL

Langflow: MCP stdio command injection allows host RCE

CVE-2026-105697
9.9
EPSS 0.9%
Code ExecutionAuth BypassFrameworkAgentPlugin
lfxPatch: 1.10.3CWE-786 ATLAS1 incident
CRITICALEXPLOIT AVAIL

vm2: node:sqlite sandbox escape allows native code RCE

CVE-2026-92938
9.9
EPSS 0.6%
Code ExecutionSupply ChainPluginAgentFramework
vm2Patch: 3.11.7CWE-6933.0K3 ATLAS
CRITICALEXPLOIT AVAIL

vm2: native code exec via crypto.setEngine sandbox escape

CVE-2026-92939
9.9
EPSS 0.6%
Code ExecutionSupply ChainPluginAgentFramework
vm2Patch: 3.11.7CWE-1143.0K3 ATLAS
CRITICALEXPLOIT AVAIL

Decepticon: ChatML token injection yields sandbox RCE

CVE-2026-61732
10.0
EPSS 1.5%
Prompt InjectionCode ExecutionAgentFrameworkInference
decepticon-sdkPatch: 1.1.17CWE-74946 ATLAS
CRITICAL

MCP Atlassian: unauthenticated access to operator Jira/Confluence

CVE-2026-77244
10.0
EPSS 0.3%
Auth BypassData ExtractionAgentPluginAPI
mcp-atlassianPatch: 0.22.0CWE-28725 ATLAS
CRITICAL

LMDeploy: pickle deserialization RCE in DistServe

CVE-2025-66455
9.8
EPSS 1.3%
Code ExecutionAuth BypassInferenceAPIFramework
lmdeployPatch: 0.16.0CWE-50214 ATLAS
CRITICALEXPLOIT AVAIL

LMDeploy: unsafe pickle deserialization allows RCE

CVE-2025-59953
9.8
EPSS 0.9%
Code ExecutionSupply ChainInferenceFramework
lmdeployPatch: 0.10.2CWE-50213 ATLAS
CRITICALEXPLOIT AVAIL

NLTK: pickle allowlist bypass enables RCE

CVE-2026-79657
9.8
EPSS 1.3%
Code ExecutionSupply ChainFrameworkModel
nltkPatch: 3.10.3CWE-50226.1K4 ATLAS
CRITICALEXPLOIT AVAIL

unstructured: SSRF via url= param exposes internal APIs

CVE-2026-71428
9.3
EPSS 0.4%
Data LeakageAuth BypassDoSFrameworkRAGAgent
unstructuredPatch: 0.24.0CWE-6013 ATLAS
CRITICALEXPLOIT AVAIL

Pimcore: DataObject field-name injection → RCE

CVE-2026-55634
9.9
EPSS 0.7%
Code ExecutionAuth BypassFramework
pimcore/pimcorePatch: 12.3.10CWE-895143 ATLAS
CRITICALEXPLOIT AVAIL

Pimcore: PHP deserialization flaw enables RCE

CVE-2026-55220
--
EPSS 0.7%
Code ExecutionFramework
pimcore/pimcorePatch: 2026.1.6CWE-5025141 ATLAS
CRITICALEXPLOIT AVAIL

LiteLLM: unauth SSTI in /prompts/test enables RCE

CVE-2026-37004
9.8
EPSS 0.8%
Code ExecutionData LeakageInferenceAPI
litellmPatch: 1.83.7CWE-133663 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