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

praisonai: MCP HTTP server unauthenticated tool invocation

GHSA-j4f3-55x4-r6q2
9.8
Auth BypassCode ExecutionData ExtractionAgentFrameworkAPI
praisonaiPatch: 1.7.2CWE-30616 ATLAS
CRITICAL

praisonai npm: AgentOS missing auth enables agent abuse

GHSA-9752-mhqh-h34f
9.4
Auth BypassData ExtractionCode ExecutionAgentFrameworkAPI
praisonaiPatch: 1.7.2CWE-30616 ATLAS
CRITICAL

praisonai: sandbox escape in codeMode → full host RCE

GHSA-p69m-4f92-2v84
9.8
Code ExecutionPrompt InjectionData ExtractionAgentFrameworkPlugin
praisonaiPatch: 1.7.2CWE-9416 ATLAS
CRITICAL

praisonai: sandbox escape gives RCE via codeMode tool

GHSA-vmmj-pfw7-fjwp
9.9
Code ExecutionSupply ChainPrompt InjectionAgentPlugin
praisonaiPatch: 1.7.2CWE-18415 ATLAS
CRITICAL

PraisonAI: unauthenticated RCE via MCP connect endpoint

GHSA-p75f-6fp4-p57w
9.8
Code ExecutionAuth BypassAgentFrameworkPlugin
praisonaiPatch: 4.6.59CWE-7815 ATLAS
CRITICAL

PraisonAI AgentOS: unauth remote agent invocation (CVSS 9.8)

GHSA-892r-p3jq-jp24
9.8
Auth BypassCode ExecutionData ExtractionAgentFrameworkAPI
praisonaiPatch: 4.6.59CWE-20017 ATLAS
CRITICAL

praisonaiagents: unauth AgentTeam API allows agent takeover

GHSA-x8cv-xmq7-p8xp
9.8
Auth BypassCode ExecutionDoSAgentFramework
praisonaiagentsPatch: 1.6.59CWE-306205 ATLAS
CRITICAL

praisonai: unauthenticated Jobs API enables remote agent RCE

GHSA-fq2m-6wqh-x44g
9.8
Auth BypassCode ExecutionData ExtractionAgentFrameworkAPI
praisonaiPatch: 4.6.59CWE-9415 ATLAS
CRITICAL

praisonai: silent auth bypass exposes agent exec endpoints

GHSA-j4hj-7hfh-g2f4
9.8
Auth BypassCode ExecutionAgentFramework
praisonaiPatch: 4.6.59CWE-30614 ATLAS
CRITICAL

PraisonAI: Unauthenticated RCE via Jobs API auth bypass

GHSA-4869-x4pr-q22x
9.8
Auth BypassCode ExecutionPrompt InjectionAgentFrameworkPlugin
praisonaiagentsPatch: 1.6.59CWE-306205 ATLAS
CRITICAL

praisonaiagents: unauthenticated MCP SSE server enables RCE

GHSA-x227-pf99-vffg
9.8
Auth BypassCode ExecutionAgentFramework
praisonaiagentsPatch: 1.6.59CWE-306205 ATLAS
CRITICALEXPLOIT AVAILSCANNER

Langflow: unauthenticated upload → DoS + path disclosure

CVE-2026-55450
9.3
EPSS 1.2%
Auth BypassDoSData LeakageFrameworkAgent
langflowPatch: 1.9.1CWE-4005 ATLAS
CRITICALSCANNER

LiteLLM: auth bypass via Host header spoofing

CVE-2026-49468
8.1
EPSS 3.0%
Auth BypassData ExtractionInferenceAPI
litellmPatch: 1.84.0CWE-29064 ATLAS
CRITICAL

backpropagate: auth bypass exposes LLM training plane

CVE-2026-48797
--
EPSS 0.3%
Auth BypassData ExtractionDoSTraining DataModelFramework
backpropagatePatch: 1.2.0CWE-3582.8K6 ATLAS
CRITICALEXPLOIT AVAIL

vllm: auth bypass exposes OpenAI inference API

CVE-2026-48746
9.1
EPSS 1.2%
Auth BypassDoSData ExtractionInferenceAPI
vllmPatch: 0.22.0CWE-444946 ATLAS
CRITICALEXPLOIT AVAIL

Langflow: unauthenticated RCE via Shareable Playground

CVE-2026-48519
9.6
EPSS 0.8%
Code ExecutionAuth BypassFrameworkAgent
langflowPatch: 1.9.2CWE-945 ATLAS
CRITICALEXPLOIT AVAIL

Jupyter Enterprise Gateway: root privilege bypass in Kubernetes

CVE-2026-44180
9.8
EPSS 0.7%
Auth BypassCode ExecutionFrameworkInference
jupyter_enterprise_gatewayPatch: 3.3.0CWE-201.9K4 ATLAS
CRITICALEXPLOIT AVAIL

transformers: trust_remote_code bypass enables RCE via model load

CVE-2026-5241
9.6
EPSS 0.6%
Supply ChainCode ExecutionFrameworkModel
transformersPatch: 5.5.0CWE-8299.8K5 ATLAS
CRITICAL

OpenMed: RCE via trust_remote_code model loading

CVE-2026-47117
9.8
EPSS 0.9%
Code ExecutionSupply ChainModelInferenceFramework
openmedPatch: 1.5.2CWE-949.8K6 ATLAS
CRITICALEXPLOIT AVAIL

praisonai-platform: member can escalate to workspace owner

CVE-2026-47413
9.6
EPSS 0.3%
Auth BypassData ExtractionAgentFramework
praisonai-platformPatch: 0.1.4CWE-26915 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