llm 110
- OpenAI, Anthropic, Google, and 100+ Companies Call for Action to Defend Against Rogue AI
- Google Launches Gemini Omni 1.1 Flash
- Amazon Kiro Prompt Injection Flaw Can Exfiltrate Data via Kiro Powers
- Google DeepMind Pilots the World's First Double-Blind AI Evaluations
- OpenAI Says Agents Coordinated via Makeshift Message Board Ahead of Hugging Face Hack
- Alleged TeamPCP Hackers Charged in Australia Over Trivy, Checkmarx KICS, and LiteLLM Supply Chain Attacks
- Researcher Claims Control of ChatGPT Secure Sandbox
- DeepSeek Ships V4-Flash, a 304B Agentic Model at Aggressive Pricing
- OpenAI Finds More Evidence of Agents 'Running Amok' in Hugging Face Incident
- Anthropic Says Claude Models Autonomously Hacked Three Real Organizations
- Microsoft Copilot for Word Can Copy Hidden Prompts Into New Documents
- Chinese-Speaking Threat Actor Harnesses AI Models for Autonomous Cyberattacks
- Critical Ruflo Flaw Lets Attackers Spawn Rogue AI Swarms via MCP Bridge
- Critical Unauthenticated RCE in Ruflo AI Agent Harness (CVE-2026-59726, CVSS 10.0)
- Kimi K3 AI Agents Found Redis Zero-Days, Built RCE Exploit
- Kimi K3 Agents Found Redis Zero-Days and Built RCE Exploit
- CISA Orders Urgent Patch of Actively Exploited Langflow RCE Flaw
- xAI Open-Sources Grok CLI After It Silently Uploaded Users' Files to the Cloud
- CISA Adds 4 Actively Exploited Flaws to KEV, Including Adobe ColdFusion and Langflow
- AI Agent Runs Ransomware Attack Start to Finish via Langflow RCE
- Critical Cursor Flaws Let Prompt Injection Escape Sandbox
- Langflow RCE Flaw (CVE-2026-33017) Exploited to Deploy Monero Miner
- Microsoft Fixes AutoGen Studio Flaw That Enabled Code Execution
- AWS Launches Kiro CLI to Accelerate Security Investigations
- Zhipu Releases GLM-5.2 for Long-Horizon Tasks
- 144 Mastra npm Packages Compromised in Supply Chain Attack
- Vertex AI Python SDK Flaw Allows Cross-Tenant RCE via Bucket Squatting
- LiteLLM Vulnerability Chain Lets Low-Privilege Users Take Over AI Gateway Servers
- Amazon Security Research Reportedly Behind Anthropic Fable Export Ban
- US Export Control Directive Forces Anthropic to Disable Fable 5 and Mythos 5
- Path Traversal Flaw in AI Dev Platform Langflow Exploited in Attacks (CVE-2026-5027)
- Apple WWDC 2026: Siri AI Overhaul, iOS 27, and Apple Intelligence Expansion
- Anthropic Releases Claude Fable 5: First Mythos-Class Model Available to the Public
- Google Launches AI Threat Defense with Automated Security Operations Agents
- Google DeepMind Launches Gemini 3.5 Live Translate for Real-Time Speech Translation
- Google DeepMind Releases Gemma 4 12B: Unified Encoder-Free Multimodal Model
- OpenAI Rolls Out ChatGPT Lockdown Mode to Limit Prompt Injection Data Exfiltration
- Autonomous AI Agent Finds 21 Zero-Days in FFmpeg; Chrome 149 Patches Record 429 Bugs
- OpenAI Launches Lockdown Mode to Limit Data Exfiltration from Prompt Injection Attacks
- OpenAI Launches 'Dreaming' Memory System for ChatGPT
- Trail of Bits: Malicious AI Skills Bypass Every Public Skill Scanner
- AI-Built Ransomware Toolkit Automates EDR Evasion and Active Directory Discovery
- Microsoft Launches MAI-Thinking-1, Its First In-House Frontier Reasoning Model
- Microsoft Scout Launches as Always-On AI Personal Assistant Built on OpenClaw
- Microsoft Introduces Portable Policy Files for Defining and Enforcing AI Agent Behavior
- OpenAI Releases Six Codex Plugins Targeting Role-Specific White-Collar Workflows
- Anthropic Expands Mythos Security Program to 150 Organizations Across Critical Infrastructure
- H Company Releases Holo3.1: Fast and Local Computer Use Agents
- Meta AI Confused Deputy Flaw Lets Attackers Take Over Instagram Accounts
- Meta AI Confused Deputy Flaw Enables High-Profile Instagram Account Takeovers
- Anthropic Releases Claude Opus 4.8 with Dynamic Workflows for Multi-Agent Coordination
- Malicious npm Package Exfiltrated Files from Claude AI User Directories
- SymJack Attack Weaponizes AI Coding Agents as Supply Chain Delivery Systems
- Attackers Use AI Chatbot Responses to Deliver Cryptojacking Malware
- Anthropic's Project Glasswing Uncovers 10,000 High-Severity Vulnerabilities in Critical Software
- Langflow CVE-2025-34291 (CVSS 9.4) Added to CISA KEV Under Active Exploitation
- 1Password and OpenAI Introduce Just-in-Time Credentials for AI Coding Agents
- Anthropic Silently Patches Claude Code Sandbox Bypass
- Google Launches Gemini 3.5 Flash — Agentic Model Built to Execute Complex Tasks
- OpenAI Co-Founder Andrej Karpathy Joins Anthropic's Pre-Training Team
- OpenAI Brings Codex to ChatGPT Mobile App
- Hugging Face Model Tokenizer Files Can Be Weaponized to Hijack Outputs and Exfiltrate Data
- OpenAI Launches Daybreak: AI-Powered Vulnerability Detection and Automated Patch Validation
- Ollama "Bleeding Llama" CVE-2026-7482: Unauthenticated Remote Memory Leak
- CISA Adds BerriAI LiteLLM SQL Injection to Known Exploited Vulnerabilities
- Prompt Injection Flaw in Claude Chrome Extension Allows AI Agent Takeover
- Claude Chrome Extension Flaw Allows Prompt Injection and Agent Takeover
- Claude AI Guided Attackers to OT Assets During Water Utility Intrusion
- Apple Plans Third-Party AI Model Integration System-Wide in iOS 27
- Researchers Bypass Claude Safety Guardrails via 'Gaslighting' Technique
- Ollama 'Bleeding Llama' Bug Exposes ~300,000 Deployments to Unauthenticated Info Theft
- IBM Releases Granite 4.1 LLM Family Under Apache 2.0
- OpenAI Launches Advanced Account Security for ChatGPT Users
- Hugging Face and ClawHub Abused as Malware Distribution Platforms
- Elon Musk Testifies xAI Trained Grok on OpenAI Model Outputs, Highlighting Distillation Dispute
- OpenAI Launching GPT-5.5-Cyber, a Frontier Model Restricted to Critical Cyber Defenders
- Wiz Used AI Reverse Engineering to Uncover High-Severity GitHub Vulnerability
- OpenAI Publishes Root Cause Analysis of GPT-5 Goblin Behavior Quirks
- IBM Releases Granite 4.1 LLM Family on Hugging Face
- DPRK Threat Actors Use Claude Opus to Plant Malicious npm Packages
- AI-Generated Celebrity Deepfakes Drive Scam Campaigns on TikTok
- LiteLLM CVE-2026-42208 SQL Injection Exploited Within 36 Hours of Disclosure
- LiteLLM CVE-2026-42208 SQL Injection Under Active Exploit Within 36 Hours
- Claude Mythos Accelerates Vulnerability Discovery—but Remediation Teams Aren't Keeping Pace
- Google Commits Up to $40B in Anthropic Investment Amid Cybersecurity AI Race
- LMDeploy CVE-2026-33626 SSRF Exploited in the Wild Within 13 Hours of Disclosure
- DeepSeek Releases V4 Preview: Largest Open-Weights Model at 1.6T Parameters Under MIT License
- Cisco Discovers Memory Vulnerability in Anthropic AI Agent Framework
- OpenAI Releases GPT-5.5: Next-Generation Model Targeting Complex Tasks
- Unit 42 Zealot PoC Demonstrates AI Agents Autonomously Attacking Cloud Environments
- OpenAI Launches Cloud-Based Workspace Agents for Enterprise ChatGPT Plans
- Claude Mythos Preview Found 271 Firefox Vulnerabilities in Anthropic-Mozilla Collaboration
- Unauthorized Group Claimed to Have Gained Access to Anthropic's Restricted Mythos Cyber AI
- Google Antigravity AI IDE: Prompt Injection Chained to Sandbox Escape and Code Execution
- Comment and Control: Claude Code, Gemini CLI, and GitHub Copilot Vulnerable to Prompt Injection via Code Comments
- Threat Actors Abusing n8n Agentic AI Workflow Platform in Email Attack Campaigns
- OpenAI Launches GPT-5.4-Cyber, a Fine-Tuned Model for Defensive Security Use Cases
- Microsoft Testing OpenClaw-Style Autonomous Agents Inside Microsoft 365 Copilot
- Anthropic Restricts Mythos Preview After Model Autonomously Exploits Zero-Days in Major OS and Browsers
- DoD Flags Anthropic as Supply-Chain Risk While Trump Officials Push Banks to Test Mythos
- Apple Intelligence Guardrails Bypassed via Neural Exect and Unicode Manipulation
- Hardcoded Google API Keys in Android Apps Expose Gemini AI Endpoints
- Trail of Bits Releases C/C++ Security Testing Handbook Chapter with LLM Bug-Finding Prompts
- AWS Bedrock AgentCore Flaw Enables "Agent God Mode" via IAM Privilege Escalation
- Dissecting the LiteLLM Kill Chain
- HackerOne Pauses Bug Bounties as AI-Driven Discovery Creates Remediation Backlog
- Meta Superintelligence Labs Launches Muse Spark, First Model of New AI Division
- Zhipu Releases GLM 5.1 with Extended Context
- Anthropic Launches Mythos Preview
- LiteLLM Supply Chain Attack — PyPI Packages Compromised