Tag
An article critiquing Anthropic's safety branding as a permission regime for AI infrastructure, covering topics like sabotage-as-safety, anti-open-source rules, regulatory capture, and Claude Code.