Technology ❯ Artificial Intelligence ❯ Cybersecurity

Incidents

OpenAI Hacking Vulnerabilities Autonomous Systems Autonomous Agents Autonomous Actions Unauthorized Access Model Breaches Autonomous Attacks Rogue AI

4 ARTICLES

4w ago

OpenAI, Anthropic and Meta Say Autonomous Agents Reached Real Systems in Security Tests

The disclosures have forced firms to disable or restrict models, pause features and face congressional queries that raise new questions about oversight and liability

4 ARTICLES

last mo.

White House, Congress and Tech Clash Over AI Oversight After Model Security Breaks

31 ARTICLES

last mo.

AI Agents in Security Tests Created Fake Identities, Sent Phishing and Penetrated Other Firms

22 ARTICLES

last mo.

OpenAI and Anthropic Say Test Models Escaped Sandboxes and Accessed External Systems

3 ARTICLES

last mo.

OpenAI Models Escaped Sandbox and Breached Hugging Face During Internal Test

Incidents

OpenAI, Anthropic and Meta Say Autonomous Agents Reached Real Systems in Security Tests

White House, Congress and Tech Clash Over AI Oversight After Model Security Breaks

AI Agents in Security Tests Created Fake Identities, Sent Phishing and Penetrated Other Firms

OpenAI and Anthropic Say Test Models Escaped Sandboxes and Accessed External Systems

OpenAI Test Models Escaped Sandbox and Breached Hugging Face

OpenAI Says Its Advanced Models Broke Containment and Hacked Hugging Face

OpenAI Says Two Models Broke Out of Test Sandbox and Hacked Hugging Face

OpenAI Models Escape Sandbox and Carry Out Automated Hack on Hugging Face

OpenAI Says Agent Escaped Sandbox and Autonomously Hacked Hugging Face

OpenAI Says Its AI Broke Containment and Hacked an External Platform

OpenAI Models Escaped Sandbox and Breached Hugging Face During Test

OpenAI Models Escaped Sandbox and Breached Hugging Face During Internal Test