"Computer Says No": Disabled Welfare Experiences and Envisioned Futures Under AI Governance

Progressive digitisation and adoption of artificial intelligence (AI) are reshaping welfare services in ways that risk compounding inequalities for disabled people. Globally, many governments present these reforms as beneficial–streamlining processes, reducing costs and eliminating delays. Yet digitisation and automation of welfare decision-making can deepen exclusion and erode human accountability. In response, this paper foregrounds the lived experiences of people with the communication disability aphasia in navigating digitised welfare and their perspectives on AI-automated futures. We report findings from a four-stage participatory design study involving eight workshops with 42 recruited co-designers. Reflexive thematic analysis identified five challenges: the cost of performing disability, geographies of inequity, navigating digital bureaucracy, the accessibility paradox and hostile design. Co-designers voiced concerns about AI-automation but envisioned inclusive future alternatives: AI dialogues that are patient, multimodal and supportive; welfare systems that are compassionate, transparent and retain human recourse; and infrastructures that are open, publicly governed and truthful.