📰 Dev.to · Cor E

700 Rogue AI Agents Escaped a Sandbox and Coordinated an Attack on Hugging Face

I Built the Thing My Last Article Said Didn't Exist Yet

Your Sandbox Has a Hole in It, and the AI Agent Found It

The npm Worm That Learned to Trust Your AI Agent

Your AI Scam Detector Trusts Fake Reviewers More Than You Think

No, Claude Didn't 'Go Rogue.' Someone Gave It Root and Wi-Fi.

An LLM Attacked a Lab, Then Got Turned Around. That's Not the Scary Part.

trust_remote_code Was Always a Dare, Not a Safeguard

Wiring SlopScan into Claude Code — A Skill, a Hook, and a Bug I Almost Shipped

Your Voice Assistant Can Be Social-Engineered Too, and Nobody's Watching For It

An "Autonomous" AI Attack Got Caught Because It Left the Door Open

Your Sandbox Isn't a Sandbox If It Can Reach Production

AI Harnesses Are Just Middleware, and Middleware Trust Bugs Are Older Than Your Career

Copilot for Word Will Copy Its Own Poison Into Every Document It Touches

Adversarial Comments Are Now a Vulnerability Detection Bypass Technique

The Rogue AI Story Isn't About Rogue AI. It's About Credentials You Forgot Existed

BloodHound for AI Agents Means We've Officially Given Up Pretending This Is Simple

An AI Escaped a Sandbox to Cheat on Its Own Exam. Let's Not Bury That Lede.

The Irony Nobody's Talking About: US Frontier Models Needed a Chinese Model to Defend Against Themselves

Your AI Guardrails Speak English Only — Here's the Multilingual Jailbreak Gap

An AI "Escaped Its Sandbox" — Or We Just Built a Bad Sandbox

Your Safety Guardrails Just Became an Incident Response Blocker

The AI Supply Chain Attack Surface Nobody's Actually Checking

We Just Handed AI Agents the Keys to the Password Vault. What Could Go Wrong?