Invisible Unicode characters popularized for hiding instructions from AI models are now being used to obfuscate words before email filters parse them.
Tag: AI Security
LLMjacking Attack Uses Leaked AWS IAM Key to Steal Paid AI Model Access
A newly documented cloud intrusion shows how quickly attackers can turn a single leaked AWS credential into a revenue stream by hijacking access to…
OpenAI Investigates Elevated Errors Affecting ChatGPT and Codex Services
OpenAI has reported an ongoing service incident involving elevated errors across ChatGPT and Codex, with the company confirming that it is investigating…
Breaking| MAJOR GLOBAL AI OUTAGE: ChatGPT, Claude, and Grok Down
A massive, widespread AI outage has struck the world’s leading generative platforms, leaving hundreds of thousands of users…
Meta releases Muse Spark 1.3 AI model
Meta Platforms has released Muse Spark 1.3, the latest version of its AI model focused on coding and agentic task performance.
Google’s Gemini 3.8 Flash takes on bigger AI models at a lower cost
Google has introduced Gemini 3.8 Flash, available to developers today, and a gated sibling, Gemini 3.8 Flash Cyber, reserved for vetted security teams.…
HiddenLayer Raises $100 Million for AI Runtime Security
The Austin-based company will invest in agentic runtime security capabilities to secure AI coding agents.
AI Agent Firewall Startup AIR Security Emerges From Stealth With $50 Million
The startup’s firewall evaluates AI skills, plugins and MCP servers for malicious instructions, excessive permissions and software supply chain risks.
Google, Anthropic, and OpenAI Unveil Advanced Cyber AI Models, Defense Programs, and Safety Controls
AI is quickly changing the way we approach cybersecurity. Google, Anthropic, and OpenAI are now developing new Cyber…
This Is Flock’s AI Search Tool for Cops
WIRED rebuilt Flock’s latest search tool from code the company sends to a police officer’s browser. Its AI can keep watch across multiple cameras for…
Attackers Expose Ongoing AI Tool Use Targeting Organizations in Latin America
Explore how attackers targeting Latin American entities use AI for data exfiltration and how basic OpSec errors allow defenders to disrupt operations.
Tencent User Data Fuels High-Performance Hy4 Preview AI Model
Hy4 preview from Tencent becomes latest open-weight model to challenge US leaders, drawing on company’s wide-ranging user data
Your AI agent’s system prompt is not a security control
An AI agent told in its system prompt to show a user only what that user is cleared to see will hand over more the moment someone talks it into doing so.…
UK Peers Propose ‘Kill Switch’ For Powerful AI Models
Group of peers in House of Lords argue government should have ability to disable powerful models to protect national security
To keep the AI hacking genie bottled up, try one-way networks
Sandboxes, permissions, and VMs aren’t enough to keep frontier models at bay. “Data diodes” might do the job
Claude AI Can Now Control macOS and Windows Computers to Click, Type and Open Apps
Anthropic has enhanced Claude’s desktop automation capabilities, enabling the AI assistant to operate directly on macOS and Windows computers through…
When AI quietly breaks things, who pays?
David Halbreich, an insurance recovery partner at Reed Smith, breaks down how AI companies should handle coverage gaps that come up as the industry grows.…
Claude AI Now Controls Your macOS and Windows Computer in the Background
Anthropic has quietly pushed one of its most consequential agentic upgrades yet, letting Claude take control of a user’s desktop and complete tasks while…
CrowdStrike FalCon 2026: AI Is Changing the Speed of Cybersecurity
CrowdStrike FalCon 2026 showed how autonomous AI is forcing cybersecurity to move faster.
CrowdStrike and NVIDIA Launch Dual-Use Cybersecurity AI
CrowdStrike SafeMind pairs offensive and defensive AI models in a closed loop designed to test attack paths and strengthen enterprise cyber defenses.