Emerging Consensus on ‘Ethical AI’: Human Rights Critique of Stakeholder Guidelines

Voluntary guidelines on ‘ethical practices’ have been the response by stakeholders to address growing concern over harmful social consequences of artificial intelligence and digital technologies. Issued dozens actors from industry, government professional associations, are creating a consensus core standards principles for ethical design, development deployment (AI). Using human rights (equality, participation accountability) attention right privacy, this paper reviews 15 preselected be strongest rights, global health. We find about half these ground their in international law incorporate key principles; even could go further, especially suggesting ways operationalize them. Those that adopt ethics framework particularly weak laying out accountability, often focusing ‘transparency’, remaining silent enforceability which would effectively protect good. These mention as rhetorical device obscure absence enforceable accountability measures, give single privacy. ‘ethics’ guidelines, disproportionately corporations other interest groups, also addressing inequalities discrimination. argue voluntary set de facto norms re-interpretation term ‘human rights’ what considered ‘ethical’ practice field. This exposes an urgent need action governments civil society develop more rigorous regulatory grounded frameworks, capable holding Big Tech powerful account.