Claude Opus Gets a Serious Safety Upgrade
Anthropic just gave Claude Opus a level-up in AI safety—think of it like locking the vault and throwing away the map. As of now, it's running on AI Safety Level 3, with tighter security to keep its model weights from being stolen and smart safeguards that block any help with bio or chem weapons. It's like installing a firewall with a conscience. One step closer to AI with built-in ethics.