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
3,023

AI/ML CVEs Tracked

424

Critical

115

New This Week

19

In CISA KEV

Sort: Date CVSS EPSS KEV first Most exploited
Filter: All Critical High Medium KEV only Active exploitation Has patch No patch

Latest AI Security Threats

Showing 20 of 3023 results
Severity CVE ID Summary CVSS EPSS Package Date
HIGH CVE-2026-31222 snorkel: RCE via insecure model checkpoint loading 8.8 0.4% snorkel May 12 HIGH CVE-2026-31221 pytorch-lightning: RCE via insecure checkpoint deserialization 7.8 0.4% pytorch-lightning May 12 UNKN CVE-2026-31219 optimate: RCE via unsafe torch.load() deserialization — 0.6% — May 12 UNKN CVE-2026-31218 optimate: unsafe torch.load() enables RCE via model file — 0.6% — May 12 CRIT CVE-2026-31214 torch-checkpoint: unsafe pickle deserialization RCE 9.8 0.5% — May 12 HIGH E CVE-2026-2614 MLflow: path traversal allows unauthenticated file read 7.5 0.7% mlflow May 11 MEDI E CVE-2026-43979 local-deep-research: HTML injection enables SSRF via WeasyPrint 5.0 0.3% local-deep-research May 11 CRIT E CVE-2026-43995 Flowise: SSRF in agent tools bypasses security wrapper 9.8 0.4% flowise May 11 HIGH E CVE-2026-2393 MLflow: SSRF in webhook URL enables cloud credential theft 7.1 0.3% mlflow May 11 HIGH CVE-2026-45224 Crabbox: path traversal enables arbitrary file wipe 7.1 0.1% crabbox May 11 HIGH E CVE-2026-45223 Crabbox: coordinator auth bypass via forged admin claim 8.8 0.4% crabbox May 11 HIGH CVE-2026-31253 flash-attention: RCE via unsafe checkpoint deserialization 7.3 0.2% flash_attn May 11 UNKN CVE-2026-31252 CosyVoice: RCE via unsafe torch.load() deserialization — 0.1% — May 11 UNKN CVE-2026-31251 CosyVoice: RCE via unsafe torch.load() deserialization — 0.2% — May 11 UNKN CVE-2026-31250 CosyVoice: RCE via unsafe torch.load() in model averaging — 0.2% — May 11 UNKN CVE-2026-31249 CosyVoice: insecure deserialization RCE via .pt files — 0.2% — May 11 HIGH E CVE-2026-44346 BentoML: Dockerfile injection enables build-time RCE 8.8 0.3% bentoml May 11 HIGH E CVE-2026-44345 BentoML: unsanitized base_image allows Dockerfile RCE 8.8 0.3% bentoml May 11 HIGH E CVE-2026-44570 open-webui: IDOR exposes cross-user AI memory data 8.3 0.3% open-webui May 11 MEDI E CVE-2026-44571 open-webui: auth bypass allows message tampering 6.5 0.3% open-webui May 11

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