Anthropic releases Mythos model, claiming its safety safeguards are now functioning effectively · Digg