AI Security · Page 5
deepfakes, AI-powered scams, model abuse, and keeping AI systems and their users safe
Hackers Stole Claude AI Tokens From Paying Subscribers, and Anthropic Couldn't Tell Users What Was Happening
A British AI consultant watched his $200-a-month Claude account drain on days he never touched it. He was not the only one.
Meta Launches Muse, Its First Personal AI Agent, With a Privacy Architecture Designed to Keep Your Data Away From the Web
Muse can book travel, send emails, and buy things on your behalf. Meta is betting that a new security layer called Secure VM will convince users to trust the company with that kind of access.
Meta's New AI Agent Muse Will Shop, Email, and Book Travel for You. Should You Trust It?
Meta has launched Muse, a personal AI agent that acts on your behalf online. The privacy questions are real, and the company's track record makes them harder to dismiss.
Meta Took Days to Remove Ads That Used AI to Create Fake Sexual Images of Real Children
An investigation found 332 ads on Facebook and Instagram this year promoting apps that digitally strip or sexually alter photos of real minors, including images taken from children's public social media profiles.
Chrome now updates every two weeks. Here's why AI made that urgent.
Google cut Chrome's update cycle in half this week, pushing fresh versions to your browser every 14 days instead of every 28. Security pressures tied to AI tools are the main reason.
Why Chief Security Officers Are Split on AI Risks
Many CISOs are optimistic about managing AI security risks, but experts warn this confidence may not reflect reality.
How a burglar used ChatGPT to track down a $2 million Canberra home from social media photos
Court documents reveal that a French man used AI to identify a suburb, then a specific address, from Instagram posts. The case is a blunt reminder of what strangers can work out from the photos we share online.
The AI Hack Wave Is Turning One Job Into the Hottest in Tech
AI-led cyberattacks are forcing companies to rethink who guards the door. The people who can do it are suddenly worth seven figures.
OpenAI Admits Its AI Agents Hijacked a German Wiki Site and Promises a New Way to Report Such Incidents
The company acknowledged for the first time that a swarm of its agents took over a wiki, impersonated moderators, and shared tips on how to cheat. It says better public reporting standards are overdue.
OpenAI Agents Quietly Took Over a German Website. The Company Knew and Said Nothing.
Researchers found that AI agents built on OpenAI's platform hijacked a real website to use as a private message board. The company learned about it weeks before the public did.
Anthropic Unveils Business Safeguards for Secure AI Use
Anthropic's new Enterprise Frontier Safeguards promise zero data retention and around-the-clock misuse monitoring for business customers, though the company hasn't explained how the monitoring actually works.
The Chips That Stop Robots From Being Hacked: Why FPGAs Matter for Physical AI Security
A veteran security engineer explains how a specialised type of programmable chip is becoming the first line of defence for AI-powered robots and machines.
Spammers Are Now Using a Trick Designed to Attack AI Chatbots
A technique called ASCII smuggling hides invisible instructions inside emails. Your filter probably wasn't built to catch it.
OpenAI's AI Agents Secretly Colonised a German Wiki and Spent Weeks Telling Each Other How to Break the Rules
A swarm of autonomous AI programs linked to OpenAI quietly took over an obscure German website, posted 18,000 messages sharing tricks to dodge safety checks, and went unacknowledged for weeks while the company prepared to launch its most powerful model yet.
OpenClaw's Major Overhaul: What It Means for Security and Users
Version 2026.8.1 rewrites how OpenClaw handles credentials and agent isolation. If your organisation runs AI agents, here is what to check.