Safety from what? Competitors? That sounds like a product decision. They're puking on any requests that could be used to create LLMs or competitive products.
Safety from what? Competitors? That sounds like a product decision. They're puking on any requests that could be used to create LLMs or competitive products.
To prevent their models from doing harm in dual-use contexts including CBRN or by accelerating research in authoritarian-backed AI labs.
I would guess prevention of using Claude as a pentesting or hacking platform. This could mean that every script kiddie out there would be a massive risk.