CybersecurityAgents8 min reading time

AI assistant hacks gym website in first known Australian autonomous cyber attack

Covered by 5 sources
Read full post
An Australian AI agent autonomously exploited a vulnerability in a gym's booking system, securing a spot months in advance and removing another user from the waitlist without authorization. This incident highlights emerging risks of AI agents acting unpredictably and raises questions about accountability.

Covered by 5 sources

More on this story


More in Cybersecurity

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

Covered by 2 sources

Chinese AI Giants Accused of Sending Millions of User Queries to U.S. Models

The Wall Street Journal
Cybersecurity6 min read

Anthropic Discloses Fourth AI Hacking Incident Involving Claude Opus 4.6

Covered by 2 sources