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: broken authz lets users overwrite channel chats

CVE-2026-59714
7.1
EPSS 0.5%
Auth BypassSocial EngineeringFrameworkAPI
open-webuiPatch: 0.10.0CWE-86222 ATLAS
LOWEXPLOIT AVAIL

onnx: OOB heap read in Gemm opset-downgrade adapter

CVE-2026-63632
3.3
EPSS 0.2%
Data LeakageDoSSupply ChainFrameworkModel
onnxPatch: 1.22.0CWE-1251.1K2 ATLAS
MEDIUMEXPLOIT AVAIL

datasets: path traversal via file_name leaks local files

CVE-2026-66007
6.5
EPSS 0.5%
Data LeakageSupply ChainTraining DataFramework
datasetsPatch: 5.0.1CWE-222.8K4 ATLAS
HIGHEXPLOIT AVAIL

Gitea: cached permission check allows repo takeover

CVE-2026-27775
8.8
EPSS 0.9%
Auth BypassSupply ChainTraining Data
code.gitea.io/giteaPatch: 1.26.3CWE-8636.3K2 ATLAS
MEDIUMEXPLOIT AVAIL

Gitea: LFS deploy-key flaw leaks private repo objects

CVE-2026-58435
5.4
EPSS 0.3%
Auth BypassData LeakageTraining DataModel
code.gitea.io/giteaPatch: 1.27.0CWE-2666.3K3 ATLAS
HIGHEXPLOIT AVAIL

Pillow: OOB read/DoS via unchecked mmap stride (McIdas)

CVE-2026-54058
--
EPSS 0.5%
Data LeakageDoSInferenceTraining DataPlugin
pillowPatch: 12.3.0CWE-1252 ATLAS
HIGHEXPLOIT AVAIL

Keras: unsafe pickle deserialization allows RCE

CVE-2026-12484
7.8
EPSS 0.3%
Code ExecutionSupply ChainFrameworkModel
kerasPatch: 3.12.3CWE-5024.1K3 ATLAS
MEDIUMEXPLOIT AVAIL

ViewComponent: stale context leaks admin UI cross-user

CVE-2026-54497
6.8
EPSS 0.3%
Auth BypassData LeakageFramework
view_componentPatch: 4.12.0CWE-362514
HIGHEXPLOIT AVAIL

yutu: arbitrary file write via MCP caption-download tool

CVE-2026-50158
7.7
EPSS 0.2%
Code ExecutionDoSPrompt InjectionAgentPlugin
github.com/eat-pray-ai/yutuPatch: 0.10.9-dev1CWE-736424 ATLAS
LOWEXPLOIT AVAIL

Keras: symlink bypass enables tar path traversal

CVE-2026-12482
3.1
EPSS 0.3%
Supply ChainCode ExecutionFrameworkModel
kerasPatch: 3.12.3CWE-224.1K4 ATLAS
HIGHEXPLOIT AVAIL

Decidim: JWT replay bypasses multi-tenant isolation

CVE-2026-45414
8.5
EPSS 0.5%
Auth BypassPrivacy ViolationData LeakageAPIFramework
decidimPatch: 0.31.5CWE-287514
CRITICALEXPLOIT AVAIL

PraisonAI: AICoder root RCE via unsanitized tool calls

CVE-2026-61445
9.9
EPSS 0.6%
Code ExecutionPrompt InjectionAgentPlugin
praisonaiPatch: 4.6.78CWE-2215 ATLAS
HIGHEXPLOIT AVAIL

PraisonAI Platform: workspace members bypass owner authz

CVE-2026-61442
7.1
EPSS 0.3%
Auth BypassAgentFramework
praisonai-platformPatch: 0.1.9CWE-86212 ATLAS
HIGHEXPLOIT AVAIL

PraisonAI: webhook signature bypass enables spoofing

CVE-2026-61428
7.3
EPSS 0.3%
Auth BypassPrompt InjectionSocial EngineeringAgentPlugin
praisonaiPatch: 4.6.78CWE-29014 ATLAS
CRITICALEXPLOIT AVAIL

PraisonAI: SQL/CQL injection via unvalidated vector dim

CVE-2026-60090
9.8
EPSS 0.7%
Code ExecutionData LeakageDoSRAGAgentFramework
praisonaiPatch: 4.6.78CWE-8913 ATLAS
MEDIUMEXPLOIT AVAIL

PraisonAI: path traversal leaks files into LLM prompts

CVE-2026-60088
5.5
EPSS 0.2%
Data LeakagePrivacy ViolationAgentFramework
praisonaiPatch: 4.6.78CWE-2213 ATLAS
CRITICALEXPLOIT AVAILSCANNER

SiYuan Note: chrome-extension origin auth bypass

CVE-2026-54069
--
EPSS 7.5%
Auth BypassData LeakageSupply ChainPluginAPI
github.com/siyuan-note/siyuan/kernelPatch: 0.0.0-20260628153353-2d5d72223df4CWE-3463.0K3 ATLAS
HIGHEXPLOIT AVAILSCANNER

SiYuan: /assets path traversal exposes API secrets

CVE-2026-54066
7.5
EPSS 2.2%
Auth BypassData LeakagePrivacy ViolationAPIPlugin
github.com/siyuan-note/siyuan/kernelPatch: 0.0.0-20260628153353-2d5d72223df4CWE-223.0K3 ATLAS
CRITICALEXPLOIT AVAIL

mcp-server-kubernetes: arg injection exfils bearer token

CVE-2026-61459
9.8
EPSS 2.4%
Auth BypassData LeakageAgentPlugin
mcp-server-kubernetesPatch: 3.9.0CWE-8814 ATLAS
HIGHEXPLOIT AVAIL

PraisonAI: unsafe dynamic import enables RCE

CVE-2026-61437
7.8
EPSS 0.2%
Supply ChainCode ExecutionAgentFramework
praisonaiagentsPatch: 1.6.78CWE-693204 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