Anthropic launches an institute to tackle AI risks - SiliconANGLE

It's responsible for studying AI -related cybersecurity risks . In one recent project, the unit used Claude to scan Firefox's code base for ...