Empowering and Centering Impacted Stakeholders in AI Design

Technology and AI risk harmful outcomes when not designed responsibly, such as the deterioration of worker well-being and inequitable resource allocation. In response, researchers have created resources for responsible AI design, from reflection aids that support practitioners' critical thinking to frameworks for eliciting stakeholder feedback about technology design. Scholars are also increasingly turning towards participatory methods to include stakeholders in design processes. However, these approaches are criticized for their ambiguity around how to incorporate reflection and participation into development and their potential for "participation washing". To address these, I propose a tool aimed at empowering impacted constituents in using their lived experiences to contextualize datasets and surface new factors for practitioners to consider. This work aims to make contributions by advancing methods for participatory AI design and surfacing initial ideas for how to integrate feedback of impacted stakeholders back into the development processes.