FrancescaG comments on Toward Safety Cases For AI Scheming