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 565 results — Medium severity, Active exploitation
MEDIUMEXPLOIT AVAIL

vLLM: token injection DoS via multimodal placeholders

CVE-2026-44222
6.5
EPSS 0.4%
DoSPrompt InjectionInferenceModelFramework
vllmPatch: 0.20.0CWE-129945 ATLAS
MEDIUMEXPLOIT AVAIL

wireshark-mcp: path traversal enables arbitrary file write via MCP

CVE-2026-43901
6.8
EPSS 0.3%
Prompt InjectionCode ExecutionSupply ChainAgentPlugin
CWE-225 ATLAS
MEDIUMEXPLOIT AVAIL

LobeChat: XSS-to-RCE via exposed Electron IPC

CVE-2026-42045
6.2
EPSS 0.3%
Code ExecutionPrompt InjectionSocial EngineeringAPIAgent
@lobehub/lobehubCWE-781.8K6 ATLAS
MEDIUMEXPLOIT AVAIL

jupyter-server: auth cookie survives password reset

CVE-2026-40934
6.8
EPSS 0.3%
Auth BypassData ExtractionFrameworkAPI
jupyter-serverPatch: 2.18.0CWE-6131.9K4 ATLAS
MEDIUMEXPLOIT AVAIL

jupyter-server: Open redirect enables credential phishing

CVE-2025-61669
--
EPSS 0.2%
Social EngineeringAuth BypassFrameworkAPI
jupyter-serverPatch: 2.18.0CWE-6011.9K5 ATLAS
MEDIUMEXPLOIT AVAIL

Langchain-Chatchat: auth bypass on file service endpoints

CVE-2026-7844
6.3
EPSS 0.3%
Auth BypassData ExtractionFrameworkRAG
CWE-2875 ATLAS
MEDIUMEXPLOIT AVAIL

SGLang: deserialization in tokenizer loader enables RCE

CVE-2026-7669
5.6
EPSS 0.4%
Supply ChainCode ExecutionFrameworkInference
sglangCWE-209.8K3 ATLAS
MEDIUMEXPLOIT AVAIL

vllm: uninitialized KV cache memory leaks inference data

CVE-2026-7141
5.6
EPSS 0.3%
Data LeakageDoSInferenceFramework
vllmPatch: 0.19.1CWE-908944 ATLAS
MEDIUMEXPLOIT AVAIL

FastChat: control flow flaw corrupts arena comparison

CVE-2026-6608
5.3
EPSS 0.3%
Adversarial ExamplesSupply ChainFrameworkAPI
fschatCWE-6707353 ATLAS
MEDIUMEXPLOIT AVAIL

Langflow: MCP config injection via X-Forwarded-For header

CVE-2026-6599
6.3
EPSS 0.2%
Auth BypassCode ExecutionSupply ChainFrameworkAgentPlugin
langflowCWE-744 ATLAS
MEDIUMEXPLOIT AVAIL

Langflow: cleartext auth storage exposes API keys

CVE-2026-6598
4.3
EPSS 0.2%
Data ExtractionPrivacy ViolationFramework
langflowPatch: 1.9.1CWE-3125 ATLAS
MEDIUMEXPLOIT AVAIL

langsmith: prototype pollution enables auth bypass, RCE

CVE-2026-40190
5.6
EPSS 0.3%
Supply ChainAuth BypassCode ExecutionFrameworkAgentPlugin
langsmithPatch: 0.5.18CWE-13213.2K4 ATLAS
MEDIUMEXPLOIT AVAIL

rembg: path traversal exposes arbitrary files via HTTP API

CVE-2026-40086
5.3
EPSS 0.6%
Data ExtractionPrivacy ViolationInferenceAPI
rembgPatch: 2.0.75CWE-221.1K4 ATLAS
MEDIUMEXPLOIT AVAIL

PraisonAI: MCP env inheritance exposes API keys

CVE-2026-40159
5.5
EPSS 0.2%
Supply ChainData LeakageData ExtractionAgentFrameworkPlugin
PraisonAIPatch: 4.5.128CWE-20016 ATLAS
MEDIUMEXPLOIT AVAIL

PraisonAI: decompression bomb causes disk exhaustion

CVE-2026-40148
6.5
EPSS 0.2%
DoSSupply ChainAgentFramework
PraisonAIPatch: 4.5.128CWE-40914 ATLAS
MEDIUMEXPLOIT AVAIL

praisonaiagents: glob traversal leaks filesystem metadata

CVE-2026-40152
5.3
EPSS 0.4%
Data ExtractionPrivacy ViolationAgentPlugin
praisonaiagentsPatch: 1.5.128CWE-22205 ATLAS
MEDIUMEXPLOIT AVAILSCANNER

PraisonAI: unauthenticated agent config and system prompt disclosure

CVE-2026-40151
5.3
EPSS 0.8%
Data ExtractionAuth BypassAgentAPI
PraisonAIPatch: 4.5.128CWE-20016 ATLAS
MEDIUMEXPLOIT AVAIL

PraisonAI: unbounded body read enables local DoS

CVE-2026-40115
6.2
EPSS 0.3%
DoSAuth BypassAgentFramework
PraisonAIPatch: 4.5.128CWE-77013 ATLAS
MEDIUMEXPLOIT AVAIL

OpenClaw: SIGKILL bypass skips security-sensitive cleanup

CVE-2026-35667
6.1
EPSS 0.1%
DoSData LeakageAgent
OpenClaw33 ATLAS1 incident
MEDIUMEXPLOIT AVAIL

OpenClaw: pre-auth resource exhaustion via Feishu webhook DoS

CVE-2026-35665
5.3
EPSS 0.3%
DoSAgent
OpenClaw33 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