Automated Decision-Making (ADM) is increasingly used in the design, delivery and governance of many social services by governments, private and not-for-profit organisations. Influential advocates (e.g. OECD 2017) argue for further automation to promote efficiencies, reduce human errors and bias, and enhance personalised service delivery, choice and resourcing.

Yet, ADM in social services is also highly problematic with detractors arguing that professionally-trained human decision making is necessary in negotiating service provision to meet complex human circumstances and needs.

There are also considerable ethical challenges relating to bias, discrimination and surveillance, particularly when used on the largely disadvantaged and marginalized populations receiving social services.