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 615 results — Active exploitation, has patch
HIGHEXPLOIT AVAIL

Open WebUI: stored XSS via citation iFrame → admin RCE

CVE-2026-26192
7.3
EPSS 0.2%
Code ExecutionData LeakageSocial EngineeringFrameworkRAG
open-webuiPatch: 0.7.0CWE-7924 ATLAS
HIGHEXPLOIT AVAIL

Open WebUI: chat XSS steals tokens, admin RCE

CVE-2025-46719
--
EPSS 0.6%
Code ExecutionAuth BypassData LeakageFrameworkPlugin
open-webuiPatch: 0.6.6CWE-7924 ATLAS
MEDIUMEXPLOIT AVAIL

Open WebUI: Stored XSS via HTML upload enables admin RCE

CVE-2025-46571
--
EPSS 0.4%
Code ExecutionSocial EngineeringData LeakageFramework
open-webuiPatch: 0.6.6CWE-7923 ATLAS
MEDIUMEXPLOIT AVAIL

onnx: crafted model triggers SIGSEGV in version_converter

CVE-2026-44512
5.5
EPSS 0.2%
DoSSupply ChainModelFramework
onnxPatch: 1.22.0CWE-4761.1K3 ATLAS
UNKNOWNEXPLOIT AVAIL

Langroid: unvalidated LLM Cypher enables graph RCE

CVE-2026-55615
--
EPSS 0.5%
Prompt InjectionCode ExecutionData LeakageAgentFrameworkRAG
langroidPatch: 0.65.5CWE-7446 ATLAS
HIGHEXPLOIT AVAIL

vLLM: crafted spec-decoding request crashes GPU worker (DoS)

CVE-2026-54234
7.5
EPSS 0.6%
DoSInference
vllmPatch: 0.24.0CWE-20942 ATLAS
MEDIUMEXPLOIT AVAIL

Coder: CORS bypass via workspace subdomain proxy

CVE-2026-55438
5.8
EPSS 0.2%
Auth BypassData LeakageTraining DataAPI
github.com/coder/coder/v2Patch: 2.34.2CWE-3466.3K2 ATLAS
HIGHEXPLOIT AVAIL

Langroid: auth bypass invokes disabled tools via raw JSON

CVE-2026-54771
8.1
EPSS 0.4%
Auth BypassCode ExecutionAgentFramework
langroidPatch: 0.65.3CWE-7443 ATLAS
CRITICALEXPLOIT AVAIL

Langroid: prompt injection to RCE via broken eval() sandbox

CVE-2026-54769
10.0
EPSS 0.9%
Prompt InjectionCode ExecutionAgentFramework
langroidPatch: 0.65.2CWE-9446 ATLAS
UNKNOWNEXPLOIT AVAIL

Langroid: SQLChatAgent regex bypass exposes pg_read_file

CVE-2026-54760
--
EPSS 0.6%
Auth BypassData ExtractionPrompt InjectionAgentFramework
langroidPatch: 0.65.1CWE-2243 ATLAS
MEDIUMEXPLOIT AVAIL

vLLM: audio upload OOMs process before size check

CVE-2026-55646
6.5
EPSS 0.5%
DoSInferenceAPI
vllmPatch: 0.24.0CWE-400942 ATLAS
CRITICALEXPLOIT AVAIL

Keras: safe_mode=None bypass enables Lambda RCE

CVE-2026-12481
9.8
EPSS 0.7%
Code ExecutionSupply ChainFrameworkModel
kerasPatch: 3.12.3CWE-5024.1K4 ATLAS
HIGHEXPLOIT AVAIL

langroid: SQL blocklist bypass leaks Postgres files

CVE-2026-50180
--
EPSS 0.7%
Prompt InjectionData ExtractionAuth BypassAgentFramework
langroidPatch: 0.64.0CWE-2244 ATLAS
HIGHEXPLOIT AVAIL

Langroid: path traversal escapes sandboxed file tools

CVE-2026-50181
7.1
EPSS 0.2%
Auth BypassData LeakageAgentFramework
langroidPatch: 0.64.0CWE-2244 ATLAS
CRITICALEXPLOIT AVAIL

mcp-memory-service: auth bypass on document API

CVE-2026-50027
9.8
EPSS 1.0%
Auth BypassData LeakageData ExtractionAPIAgentRAG
mcp-memory-servicePatch: 10.67.1CWE-3062.8K4 ATLAS
HIGHEXPLOIT AVAIL

Gradio: path traversal in FileExplorer leaks files

CVE-2026-49119
7.5
EPSS 0.7%
Data LeakagePrivacy ViolationFramework
gradioPatch: 6.16.0CWE-227353 ATLAS
HIGHEXPLOIT AVAIL

auth-fetch-mcp: SSRF bypass via IPv6 loopback

CVE-2026-49857
7.4
EPSS 0.5%
Auth BypassData LeakageAgentPlugin
auth-fetch-mcpPatch: 3.0.2CWE-9184 ATLAS
HIGHEXPLOIT AVAIL

Ray: RCE via pickle/torch deserialization in WebDataset

CVE-2026-57516
8.8
EPSS 0.9%
Supply ChainCode ExecutionTraining DataFramework
rayPatch: 2.56.0CWE-5026424 ATLAS
MEDIUMEXPLOIT AVAIL

Keras: incomplete fix reopens HDF5 arbitrary file read

CVE-2026-12480
5.5
EPSS 0.2%
Data LeakageSupply ChainPrivacy ViolationModelTraining Data
kerasPatch: 3.12.3CWE-734.1K4 ATLAS
CRITICALEXPLOIT AVAILSCANNER

Langflow: authenticated RCE enables credential theft

CVE-2026-7873
9.9
EPSS 6.2%
Code ExecutionData LeakagePrivacy ViolationFrameworkAgent
langflowPatch: 1.10.1.rc0CWE-944 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