Co-Designing AI Systems with Value-Sensitive Citizen Science


METADATA ONLY
Loading...

Date

2025-06-22

Publication Type

Working Paper

ETH Bibliography

yes

Citations

Altmetric
METADATA ONLY

Data

Rights / License

Abstract

As artificial intelligence (AI) systems increasingly shape everyday life, integrating diverse community values into their development becomes both an ethical imperative and a practical necessity. This paper introduces Value Sensitive Citizen Science (VSCS), a systematic framework combining Value Sensitive Design (VSD) principles with citizen science methods to foster meaningful public participation in AI. Addressing critical gaps in existing approaches, VSCS integrates culturally grounded participatory methods and structured cognitive scaffolding through the Participatory Value-Cognition Taxonomy (PVCT). Through iterative value-sensitive participation cycles guided by an extended scenario logic (What-if, If-then, Then-what, What-now), community members act as genuine coresearchers-identifying, translating, and operationalizing local values into concrete technical requirements. The framework also institutionalizes governance structures for ongoing oversight, adaptability, and accountability across the AI lifecycle. By explicitly bridging participatory design with algorithmic accountability, VSCS ensures that AI systems reflect evolving community priorities rather than reinforcing top-down or monocultural perspectives. Critical discussions highlight VSCS's practical implications, addressing challenges such as power dynamics, scalability, and epistemic justice. The paper concludes by outlining actionable strategies for policymakers and practitioners, alongside future research directions aimed at advancing participatory, value-driven AI development across diverse technical and sociocultural contexts.

Publication status

Editor

Book title

Journal / series

Volume

Pages / Article No.

Publisher

Social Science Research Network

Event

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Participatory AI; Algorithmic governance; Value-sensitive design; Human AI symbiosis

Organisational unit

03784 - Helbing, Dirk / Helbing, Dirk check_circle

Notes

Funding

Related publications and datasets