Anthropic makes Claude take actions on your behalf by default

Anthropic is defaulting Claude to an 'auto mode' that acts autonomously, saying it judges risk better than humans do.