Responsible AI - on whose terms?