Speaker
Description
Despite growing awareness of structural inequalities in physics, unconscious bias remains a persistent challenge in selection processes – from awards and speaker invitations to hiring. Gender disparity, institutional concentration, and career-stage imbalance often remain invisible until after decisions are made.
This talk introduces Equity Lens, a prototype developed under the Working Group Equal Opportunities (AKC) of the German Physical Society (DPG). It provides committees with a privacy-preserving dashboard of potential bias indicators before final decisions. It does not rank candidates or store individual votes – it makes hidden patterns visible. The tool calculates career-stage-adjusted metrics (m-quotient) and flags gender imbalance, institutional over-representation, and repeat invitations.
Equity Lens was presented as a selected side event at the UN Global Dialogue on AI Governance (Geneva, July 2026) [1]. The prototype is now publicly accessible as a live demo on GitHub and open for feedback [2]. The talk presents the design, feedback, and roadmap toward an open-source tool for fairer selection in physics.
[1] https://www.un.org/global-dialogue-ai-governance/sites/default/files/2026-07/13_o-20_equity_lens_ai_dialogue_concept_note_template_v3_clean1.pdf
[2] https://akc-equity-lens.github.io/equity-lens/
