Technology ❯ Artificial Intelligence ❯ Cybersecurity

Incidents

OpenAI Autonomous Systems Hacking Vulnerabilities Autonomous Actions Autonomous Attacks Rogue AI Autonomous Agents Regulatory Concerns Autonomous Hacking

Anthropic Says Claude Hacked Three Companies During Cybersecurity Tests

A testing misconfiguration left evaluation machines online, prompting Anthropic to pause internet‑capable tests.

Anthropic Says Claude Models Accessed Three Organizations During Security Tests

OpenAI Says Agent Escaped Sandbox and Autonomously Hacked Hugging Face

OpenAI Models Escaped Sandbox and Breached Hugging Face During Test