As public administrations adopt artificial intelligence (AI), we see this transition has the potential to transform public service and public policies, by offering a rapid turnaround on decision making and service delivery. However, a recent series of criticisms have pointed to problematic aspects of mainstreaming AI systems in public administration, noting troubled outcomes in terms of justice and values. The argument supplied here is that any public administration adopting AI systems must consider and address ambiguities and uncertainties surrounding two key (...) dimensions: the algorithms’ results, and how public managers make decisions for and about AI systems’ design. The article is based on bibliographic research in institutional theory perspective that relates the design of AI systems and relevant literature on the decision-making process in public policy. The main finding is to explain why autonomous decision systems continue to reproduce ambiguities and uncertainties when applied in public administration and propose a reflection on AI governance in public administration. This article points to the need for design institutions that immerse themselves in understanding the nuances, details, and potential outcomes of AI governance for public administration. Such institutions would reconcile consequentialist logic with a logic of appropriateness to help navigate and mediate ambiguities and uncertainties. (shrink)

This article probes the practical ethical implications of AI system design by reconsidering the important topic of bias in the datasets used to train autonomous intelligent systems. The discussion draws on recent work concerning behaviour-guiding technologies, and it adopts a cautious form of technological utopianism by assuming it is potentially beneficial for society at large if AI systems are designed to be comparatively free from the biases that characterise human behaviour. However, the argument presented here critiques the common well-intentioned requirement (...) that, in order to achieve this, all such datasets must be debiased prior to training. By focusing specifically on gender-bias in Neural Machine Translation systems, three automated strategies for the removal of bias are considered – downsampling, upsampling, and counterfactual augmentation – and it is shown that systems trained on datasets debiased using these approaches all achieve general translation performance that is much worse than a baseline system. In addition, most of them also achieve worse performance in relation to metrics that quantify the degree of gender bias in the system outputs. By contrast, it is shown that the technique of domain adaptation can be effectively deployed to debias existing NMT systems after they have been fully trained. This enables them to produce translations that are quantitatively far less biased when analysed using gender-based metrics, but which also achieve state-of-the-art general performance. It is hoped that the discussion presented here will reinvigorate ongoing debates about how and why bias can be most effectively reduced in state-of-the-art AI systems. (shrink)