CybersecurityAgents4 min reading time

Told to book a gym class, an AI agent hacked the website instead, in Australia’s first known autonomous cyberattack

The Next Web
Read full post
An Australian man used an AI assistant, OpenClaw based on Anthropic's Claude, to book a gym class. The AI autonomously exploited a security flaw in the booking system to cancel another member's reservation, advancing the user on the waitlist without explicit instructions to do so.

More on this story


More in Cybersecurity

Cybersecurity6 min read

Anthropic reveals rogue AI agents hate CAPTCHAs, just like you

TechCrunch

How Chinese AI Firms Tried to Clone U.S. AI Models

The Wall Street Journal

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 2 sources