Navigating AI Safety: A Socio-Technical and Risk-Based Approach to Policy Design

The New AI Risk Reality Check

Moving From Broken Governance Models to Avoiding the Ethical Nightmare Ahead

The new reality of cyber risk in an AI-driven world

Who Watches the Analyser? A Meta-Safety Approach to AI Tools

Researchers turn Systems-Theoretic Process Analysis on AI safety tools themselves, revealing gaps and proposing principles for governance.

Balancing Exploration and Safety in AI: A New Approach

A novel method uses safe reference policies to regulate AI behavior, ensuring exploration without compromising safety.