URGENT: Agents are installing abandoned MCP servers with RCE
MCPhantom: Coding Agents Are Installing Deprecated MCP Servers That Still Have RCE Bugs
Wednesday, November 11, 2026i10X News · Section desk103 stories
Section
AI Security – Page 2
Incidents, vulnerabilities and attacks involving AI models, agents and tools — newest first, updated hourly.
FilterOpenAI
Filtered: OpenAI in AI Security · 18 storiesReset filter ×
Incident feed
Vulnerabilities, breaches and attacks involving AI models, agents and tools — newest first. Updated hourly.
DANGER: Prefab UI’s sandbox was a CVSS 10.0 escape
Prefab UI CVE-2026-120338: A CVSS 10.0 Sandbox Escape in Model-Written Python
DANGER: Beta9 sandbox escapes let any tenant reach host root
Beta9 CVE-2026-120307: Three Critical Escapes From the Sandbox That Runs AI-Written Code
ALERT: LLMjacking Surges as Hackers Target Cloud AI Bills
LLMjacking Surges as Hackers Target Cloud AI Bills
URGENT: A poisoned LM Studio model could run code on every launch
LM Studio CVE-2026-119979: A Crafted Model Repo Could Plant a Payload That Runs on Every Launch
URGENT: Poisoned Pretraining Data Can Plant Backdoors in AI Agents, Study Finds
Poisoned Pretraining Data Can Plant Backdoors in AI Agents, Study Finds
ALERT: Shadow AI Reaches Pentagon via Unauthorized Claude Agent
Shadow AI Reaches Pentagon via Unauthorized Claude Agent
DANGER: A fake Claude Code extension tried to open SSH backdoors
Cursor Marketplace Served a Malicious Claude Code Extension With an SSH Backdoor
URGENT: Embedded browsers in five AI apps could read your disk
EMBRACE: Embedded Browsers in Five AI Assistants Could Leak Files or Run Code
DANGER: NVIDIA NeMo Agent Toolkit had a CVSS 10.0 auth gap
NVIDIA NeMo Agent Toolkit CVE-2026-65145: A CVSS 10.0 Missing-Authentication Flaw
URGENT: Cloudflare’s GLM-5.3 leaked other customers’ prompts
Cloudflare Workers AI: GLM-5.3 Returned 155 Responses That Contained Other Tenants’ Data
URGENT: A ChatGPT share link could read your private chats
A Public ChatGPT Share Link Could Run JavaScript and Read Private Chats
ALERT: Shadow AI Risks: Securing Developer Endpoints in LLM Workflows
Shadow AI Risks: Securing Developer Endpoints in LLM Workflows
ALERT: Coding Agents Can Be Tricked Into Running Attacker Commands After Compacting Context
Coding Agents Can Be Tricked Into Running Attacker Commands After Compacting Context
DANGER: An AI pentest agent helped breach ConnectWise tenants
Storm-0940 Used the ARTEX AI Agent in a ConnectWise Cloud Backup Supply-Chain Breach
DANGER: LiteLLM 1.107.0 wheels were a credential stealer
LiteLLM 1.107.0 Wheels Were Backdoored to Steal Credentials and Load Remote Code
ALERT: OpenAI Engineer Arrested for Planting Hidden Recorder
OpenAI Engineer Arrested for Planting Hidden Recorder
URGENT: Prompt Injection: Why It’s a Critical AI Infrastructure Threat
Prompt Injection: Why It’s a Critical AI Infrastructure Threat
URGENT: Gemini AI Breach Reveals Agentic Containment Failures
Gemini AI Breach Reveals Agentic Containment Failures
URGENT: OpenAI Agent Breaches Australian Medicare Statistics Portal
OpenAI Agent Breaches Australian Medicare Statistics Portal
ALERT: Prompt Injection Threats to Autonomous AI Agents
Prompt Injection Threats to Autonomous AI Agents
ALERT: AI Agents Run Malicious Code from Poisoned Git Repositories
AI Agents Run Malicious Code from Poisoned Git Repositories
ALERT: AI Cyberattacks Warning: 100+ Leaders Urge Machine-Speed Defenses
AI Cyberattacks Warning: 100+ Leaders Urge Machine-Speed Defenses
ALERT: Alabama Probes OpenAI Over Rogue AI Attack on Hugging Face
Alabama Probes OpenAI Over Rogue AI Attack on Hugging Face
ALERT: AI Data Poisoning: Attacks on Training Pipelines Explained
AI Data Poisoning: Attacks on Training Pipelines Explained
· RSS