Technology ❯ Artificial Intelligence ❯ AI Safety

Cybersecurity

Vulnerabilities Incident Response Model Evaluation Testing Environments Autonomous Agents Infrastructure Security Mental Health Open-Weight Models Incidents Corporate Responsibility

3 ARTICLES

2w ago

Irregular Says Human Error Let Testing AI Attack Real Systems

A naming error that matched a fictional target to a real domain exposed gaps in Irregular’s testing controls, prompting the firm to tighten evaluation protocols.

7 ARTICLES

last mo.

Cybersecurity

Irregular Says Human Error Let Testing AI Attack Real Systems

Hassabis and U.S. Officials Push FINRA-Style Vetting for Frontier AI

Germany to Create National AI Security Institute

Study Finds Poetic Prompts Can Jailbreak Major AI Chatbots

Atlas and Other AI Browsers Confront Real-World Exploits as Rollout Expands

DHS Unveils AI Framework for Securing Critical Infrastructure

U.S. Warned of 'Extinction-Level' AI Threats, Urged to Act