One Email Hijacked an AI Agent: Salt Labs' Manus Research Explained
Salt Labs showed one crafted email could make the Manus AI agent run code and reach connected accounts. How it worked and how to connect agents safely.
x47.c Malware Uses Grok to Stay Hidden and Can Burn Your AI Credits
The x47.c Windows botnet advertises an AI module that uses xAI's Grok for persistence and a mode that drains paid AI API credits. How to protect your keys.
Chrome 155 Fixes 247 Security Flaws, and AI Found Some of the Worst
Chrome 155 patches 247 security bugs, four critical. Twelve were found with help from Anthropic's Claude and two by OpenAI's Codex Security. Update now.
Did Meta's Muse AI Agent Read a Journalist's Texts? What We Know
A journalist says Meta's Muse agent read his Mac messages without permission. Meta says it can't. The claims, the evidence, and what to check on your Mac.
Apple Is Tightening macOS Full Disk Access Because of AI Agents
Apple says new Full Disk Access controls are coming to macOS because AI agents raise the stakes. What changed, why, and how to audit your Mac today.