Paper Session 5: Critical and Responsible HCI
Date: Tuesday 3 Nov 2026
Time: 14:00–14:50
Add this session to your calendar, or click a talk’s time to add just that talk.
| Time | Presentation | Abstract | Authors | Contribution type |
|---|---|---|---|---|
| 14:00-14:10 | From Accountability to Disclosure: A Citation Analysis of Positionality Statements in HCI | Positionality statements have grown rapidly in HCI, yet their theoretical grounding remains unclear. We conduct a citation analysis of ACM DL publications making claims about positionality statements, together with the literature they cite. From 36 papers we extract 48 excerpts citing 26 distinct works. Only 9 of those works engage with positionality statements as the concept is used in HCI, and only 3 offer affirmative and robust grounding. Citation practices transform reflexivity’s processual accountability into formalized disclosure, and redirect standpoint epistemology from elevating marginalized knowledge towards centring researcher identities. Positionality statements, as currently theorized, therefore risk undermining the goals of accountability and epistemic justice they seek to advance. We suggest resisting formalized disclosure, operationalizing reflexive transparency through process documentation, and re-centring empirical focus on marginalized knowledge. Concerns about the practice have so far been largely conceptual; we provide systematic citational evidence for them. | Ferreira, Sanches | Paper |
| 14:10-14:20 | Supporting Teachers in Teaching Critical AI Literacy from Co-design to Classroom Implementation | As Artificial Intelligence (AI) becomes increasingly ubiquitous in children’s everyday lives, it is crucial to develop children’s critical AI literacy. Yet, teaching critical AI literacy remains challenging in practice, with classrooms defaulting to technical skill acquisition while marginalising critical reflections on AI’s implications. Recent reviews call for supporting teachers in teaching critical AI literacy. In response, this paper explores how teachers develop strategies for teaching critical AI literacy through co-design and what happens when they implement them in the classroom. Following a design-based research approach, we trace one teacher’s journey in detail across co-design workshops, pre- and post-implementation interviews, and observation in a 7th-grade classroom. Using dialogic formative assessment as an analytical lens, we identify opportunities and challenges across four phases: co-design workshops, independent lesson planning, classroom implementation, and post-implementation reflection. We propose research directions for HCI research for designing support for facilitating critical AI literacy in formal education: (1) supporting teachers in developing dialogic facilitation repertoires; and (2) designing tools as dialogic partners that scaffold critical reflection on AI’s societal implications. | Liu, Surrow, Shokeen, Connelly, Jensen, Hjorth, Andersen, Dindler, Petersen, Iversen | Paper |
| 14:20-14:30 | Designing a Board Game to Explore and Communicate the Environmental Impact of AI | As technological developments and financial investments into AI continue, the physical footprint of these systems becomes more pronounced. The environmental impact of AI is complex and sits at the intersection of two major issues: climate change and technological innovation. Because of the scale of each issue, learned helplessness and disempowerment are profound. To address this, drawing on research through design methodologies and board games for sustainability, we developed SustAIn. SustAIn is a 2-4 player board game that enables players to move throughout a tabletop world, making decisions that affect the game state as an analogy for the environment. The game is designed around the use of narratives that communicate the environmental impact of AI. In this way, it seeks to establish a qualitative point of access to complex environmental considerations that do not always have a robust or comprehensive quantitative context. A print-and-play version of the game is available here: https://sustain-game.gitlab.io | Gillespie, Legon, O'Brien, Lopez, Barany, Lechelt | Paper |
| 14:30-14:40 | Conversational Care: Designing Responsible Human-AI Care Coordination | Healthcare systems worldwide face demand they cannot meet, and conversational artificial intelligence (CAI) is one of the pathways to scale provisions. But doing so responsibly requires addressing safety, fairness, effectiveness, accessibility, and several other criteria at once. Few projects have the resources to build a system that meets all of them from scratch. We followed a 2-year Research through Design process to iteratively build Conversational Care: a customisable platform that integrates AI agents into care-team workflows, automating routine tasks and escalating to humans when oversight is needed. The platform embeds nine responsible-AI criteria drawn from the literature. It has been designed with, and to support, community health workers, nurses, and families of people living with dementia in Peru and the UK; health coaches and link workers in London; with early exploration of care home monitoring in Singapore. Working across such varied scenarios surfaced the design trade-offs and recurring patterns that emerge when CAI is integrated into real care systems – contributions that other teams can adapt to their own contexts. | Da Re, Fonseca, Cho, Sanchez, Irina Serban, Diez-Canseco, Cuba-Fuentes, Vaidyanathan, Calvo | Paper |
| 14:40-14:50 | Class in the Prompt: Language-Based Proxy Bias in AI Systems in Bangladesh and Its Implications for the Global South | Artificial Intelligence (AI) systems are increasingly used across the Global South for consequential decisions in hiring, credit, and health triage, often before regulatory oversight is in place. This study examines a bias that standard fairness audits are not designed to detect. Large language models (LLMs) produce different outputs for identical content depending on how it is written, without naming writing style in their reasoning; where the language a person writes in signals class position, this becomes a hidden channel of discrimination. We report a matched-guise prompt audit across four frontier LLMs and three decision tasks, varying the register of 30 matched scenarios across five language varieties while holding content constant. In Bangladesh, English from Bangla-medium schooling is scored consistently lower than identical English-medium English across all three tasks and three of four models (gradient: 0.37 hiring, 0.22 loan, 0.17 triage), an effect absent from stated reasoning in 96.4% of cases. We propose a register parity check as a low-cost intervention applicable across the Global South. The findings connect to United Nations (UN) Sustainable Development Goals 10 (Reduced Inequalities) and 16 (Peace, Justice and Strong Institutions), and the governance responses proposed require no new legislation. | Ahmed, Arpita, Islam, Rahman, Zaber, Amin | Paper |