Why blocking AI models won’t stop the cyber threats they create

Anthropic forms institute to study long-term AI risks facing society