AI security Archives • Daily CyberSecurity

Puzzlemask AI Attack Vector Bypasses LLM Guardrails Diagram illustrating the Puzzlemask AI attack and Puzzlemask AI attack vector

Puzzlemask AI Attack Vector Bypasses LLM Guardrails

Do Son September 14, 2026

Check Point Research published a technical analysis, revealing a stealthy prompt injection method. The newly documented Puzzlemask...

Adversarial AI Threats Drive New Cyber Operations Adversarial AI threats mapping AI credential harvesting

Adversarial AI Threats Drive New Cyber Operations

Do Son September 14, 2026

At a Glance Actor or Group: Suspected state-sponsored groups (UNC6508, BASIN CASTLE) and cyber criminals (UNC6780). Activity...

US Agencies Warn of Malicious AI Distillation Campaigns Diagram showing malicious AI distillation campaigns targeting frontier models

US Agencies Warn of Malicious AI Distillation Campaigns

Do Son September 8, 2026

At a glance On September 9, 2026, federal security agencies exposed intellectual property theft targeting American artificial...

METR Security Breach: Stolen API Key Burns $600K in Credits METR security breach where a stolen API key consumed roughly $600K in AI credits via an exposed agent dashboard

METR security breach where a stolen API key consumed roughly $600K in AI credits via an exposed agent dashboard

METR Security Breach: Stolen API Key Burns $600K in Credits

Do Son September 2, 2026

The AI safety-evaluation organization METR, the very research body that investigated OpenAI agents attacking Hugging Face, recently...

CVE-2026-47627: CVSS 9.8 Denial of Service Hits NVIDIA Triton NVIDIA Triton vulnerability CVE-2026-47627 CVSS 9.8 denial of service advisory chart for the AI inference server

NVIDIA Triton vulnerability CVE-2026-47627 CVSS 9.8 denial of service advisory chart for the AI inference server

CVE-2026-47627: CVSS 9.8 Denial of Service Hits NVIDIA Triton

Do Son August 20, 2026

TL;DR NVIDIA patched a critical NVIDIA Triton vulnerability tracked as CVE-2026-47627. The flaw scores a CVSS base...

PoC Discloses for CVE-2026-64849: watchTowr Sees Attacks on MLflow SSRF MLflow vulnerability CVE-2026-64849 unauthenticated full-read SSRF webhook redirect bypass reaching cloud metadata

MLflow vulnerability CVE-2026-64849 unauthenticated full-read SSRF webhook redirect bypass reaching cloud metadata

PoC Discloses for CVE-2026-64849: watchTowr Sees Attacks on MLflow SSRF

Do Son August 18, 2026

TL;DR A default MLflow tracking server can be turned into a proxy for reading internal services. Tracked...

Chinese Threat Actor Runs Autonomous AI Cyberattacks Using DeepSeek JADEPUFFER agentic ransomware attack chain diagram from Langflow RCE to database extortion

JADEPUFFER agentic ransomware attack chain diagram from Langflow RCE to database extortion

Chinese Threat Actor Runs Autonomous AI Cyberattacks Using DeepSeek

Do Son August 5, 2026

At a glance Field Detail Actor Chinese-speaking operator, aliases knaithe and KnYuan Activity AI-enabled autonomous hacking campaign...

Claude Models Accessed Real Systems During Cybersecurity Evaluations Claude cybersecurity evaluations incident where an AI model reached real systems, disclosed by Anthropic

Claude cybersecurity evaluations incident where an AI model reached real systems, disclosed by Anthropic

Claude Models Accessed Real Systems During Cybersecurity Evaluations

Do Son July 31, 2026

What Anthropic disclosed Anthropic has revealed that its Claude models breached three real organizations during safety testing....

Microsoft Unveils Project Perception, an Agentic Security System Microsoft Project Perception agentic security system with red, blue, and green AI team agents, a security context layer, and actuators for the AI era

Microsoft Project Perception agentic security system with red, blue, and green AI team agents, a security context layer, and actuators for the AI era

Microsoft Unveils Project Perception, an Agentic Security System

Do Son July 28, 2026

Attackers already strike at machine speed with AI. Defenders, meanwhile, still drown in thousands of indigestible security...

Threat Actor Rebuilt a Live Botnet in Six Minutes Using an AI Coding Agent AccountDumpling Phishing Google AppSheet Abuse AI-Generated Malware PureRAT Campaign RondoDoX Botnet, Next.js React2Shell EchoGather, Paper Werewolf Salt Typhoon, Telecom Espionage BreachForums, Conor Fitzpatrick Ransomware Negotiation, DOJ Investigation MirrorFace group - Earth Kasha Emperor Dragonfly

AccountDumpling Phishing Google AppSheet Abuse AI-Generated Malware PureRAT Campaign RondoDoX Botnet, Next.js React2Shell EchoGather, Paper Werewolf Salt Typhoon, Telecom Espionage BreachForums, Conor Fitzpatrick Ransomware Negotiation, DOJ Investigation MirrorFace group - Earth Kasha Emperor Dragonfly

Threat Actor Rebuilt a Live Botnet in Six Minutes Using an AI Coding Agent

Do Son July 21, 2026

At a Glance Actor Russian-speaking solo actor using the handle “bandcampro” Activity type AI-assisted C&C botnet, password...

Mid-July 2026: Weekly Threat Intelligence Report actively exploited vulnerabilities

Mid-July 2026: Weekly Threat Intelligence Report

Do Son July 20, 2026

Welcome to the CVE WATCHTOWER briefing. Our latest weekly threat intelligence report reveals an alarming surge in...

Hugging Face Discloses Production Breach Driven by an Autonomous AI Agent Hugging Face breach driven by an autonomous AI agent attack exposing internal datasets and credentials

Hugging Face Discloses Production Breach Driven by an Autonomous AI Agent

Do Son July 18, 2026

At a glance Organization Hugging Face Data exposed A limited set of internal datasets and several service...

CrowdStrike Details New Prompt Injection Attacks on AI Diagram showing how prompt injection attacks and prompt injection techniques threaten AI security.

Diagram showing how prompt injection attacks and prompt injection techniques threaten AI security.

CrowdStrike Details New Prompt Injection Attacks on AI

Do Son July 14, 2026

TL;DR CrowdStrike added 18 new methods to its AI security taxonomy. These new prompt injection attacks target...

Phantom Squatting Attacks Weaponize AI Hallucinated Domains, Unit 42 Warns Phantom squatting attacker registering AI hallucinated domains predicted by LLM chatbots

Phantom squatting attacker registering AI hallucinated domains predicted by LLM chatbots

Phantom Squatting Attacks Weaponize AI Hallucinated Domains, Unit 42 Warns

Do Son July 8, 2026

At a glance Details Activity type Phantom squatting: registering AI hallucinated domains for phishing and malware Actors...

NVIDIA Patches Three High-Severity NeMo Framework Code Injection Flaws NVIDIA NeMo vulnerability CVE-2026-24155, CVE-2026-24252, CVE-2026-24228, code injection NVIDIA DALI vulnerabilities NVIDIA Driver Security Update CVE-2026-24187 Linux

NVIDIA NeMo vulnerability CVE-2026-24155, CVE-2026-24252, CVE-2026-24228, code injection NVIDIA DALI vulnerabilities NVIDIA Driver Security Update CVE-2026-24187 Linux

NVIDIA Patches Three High-Severity NeMo Framework Code Injection Flaws

Do Son June 17, 2026

NVIDIA has issued an urgent fix for its NeMo Framework, the popular open-source toolkit for building generative...

Critical 9.8 CVSS: NVIDIA Triton Inference Server Authentication Bypass Threatens AI Workloads NVIDIA NeMo vulnerability CVE-2026-24155, CVE-2026-24252, CVE-2026-24228, code injection NVIDIA DALI vulnerabilities NVIDIA Driver Security Update CVE-2026-24187 Linux

Critical 9.8 CVSS: NVIDIA Triton Inference Server Authentication Bypass Threatens AI Workloads

Do Son May 20, 2026

NVIDIA has released a software update for the NVIDIA Triton Inference Server to address a wave of...

Apple Google EU alliance DMA European Commission Breach Trivy Supply Chain Attack Europa.eu Breach EU Cloud Infrastructure EU Cyber Sanctions State-Sponsored Hacking EU 2040 Emissions Target, Europe Climate Leadership AWS Azure DMA Cloud Gatekeeper DSA violation, illegal content Apple DMA Delay, iPhone Mirroring EU EU Age Verification, Google Play Integrity Corning Antitrust, EU Competition Apple EU Digital Markets Act App Store commission European Union cyberattacks - InvestAI EU Targets Musk’s X Digital Markets Act, EU fines

Apple Joins Google to Fight EU Mandate on Deep AI System Access

Do Son May 15, 2026

According to recent intelligence, Apple has released an official submission to the European Commission, aligning itself with...

The “Unbreakable” OS: Android 17 Unveils AI-Powered Fraud Defense and Thief-Proof Biometrics Android 17 security features

The “Unbreakable” OS: Android 17 Unveils AI-Powered Fraud Defense and Thief-Proof Biometrics

Do Son May 13, 2026

In the face of increasingly rampant transnational financial fraud, sophisticated social engineering, and the rising tide of...

PraisonAI CVE-2026-44338 Exploited in the Wild Hours After Patch Disclosure PraisonAI Authentication Bypass Exploited in the Wild CVE-2026-44338

PraisonAI Authentication Bypass Exploited in the Wild CVE-2026-44338

PraisonAI CVE-2026-44338 Exploited in the Wild Hours After Patch Disclosure

Do Son May 13, 2026

A new report from the Sysdig Threat Research Team (TRT) reveals that on May 11, 2026, a...

9.6 Severity: Critical “Cline” AI Agent Flaw Allows Stealthy RCE via Your Browser Cline AI Vulnerability Cross-Origin WebSocket Hijacking

9.6 Severity: Critical “Cline” AI Agent Flaw Allows Stealthy RCE via Your Browser

Do Son May 12, 2026

In the rapidly evolving world of AI-assisted development, tools like Cline have become indispensable, living in editors...

AI Hijacked: Critical Claude Chrome Extension Flaw Allows Malicious Scripts to Control Your AI Claude Extension Vulnerability AI Agent Hijacking

Claude Extension Vulnerability AI Agent Hijacking

AI Hijacked: Critical Claude Chrome Extension Flaw Allows Malicious Scripts to Control Your AI

Do Son May 11, 2026

Artificial intelligence assistants are rapidly becoming integrated into our daily workflows, but what happens when a trusted...

Data Destruction and Memory Poisoning: Spring AI Vulnerabilities Threaten LLM Integrity Spring AI Vulnerability LLM Memory Poisoning Spring AI Vulnerability Vector Store Injection

Spring AI Vulnerability LLM Memory Poisoning Spring AI Vulnerability Vector Store Injection

Data Destruction and Memory Poisoning: Spring AI Vulnerabilities Threaten LLM Integrity

Do Son May 11, 2026

Spring AI, a popular framework designed to simplify AI integration for Spring developers, has issued an security...

Critical Alert 3 Active Exploits Detected Today