Britain steps up its defences against AI risks

The UK government is expanding a program that uses advanced AI models to identify vulnerabilities in its systems, having found over 1,000 issues in pilots. It is also considering legislation for AI incident reporting and managing AI systems that operate beyond human control.