Abstract
Development of policy tools and instruments has always trailed behind technological innovations. More recently, the rapid advancement of AI technologies has resulted in significant widening of this policy gap — a genuine wicked problem. To address this we developed a participatory AI Policy Design toolkit, engaging 23 technology researchers and practitioners to (1) critically evaluate existing AI use cases to identify key considerations for policy; (2) generate a catalogue of current and near future AI use-cases in their domains of expertise; and (3) collaboratively develop AI policy intervention proposal artefacts, and foster a shared vocabulary for AI policy design. Results highlight key tensions, challenges, and opportunities for collaboratively exploring AI policy discourse. The central contribution of this work is to help identify AI policy blind spots, equity concerns, and anticipate enforcement gaps, and building capacity to help democratize discourse about AI policy through contextually relevant, generative, and reflexive approaches.
Keywords
AI Policy Design, Participatory Design Tools, Wicked Problem, Governance
DOI
https://doi.org/10.21606/drs.2026.1686
Citation
Khan, A., Saghir, S., and Henman, P. (2026) The wicked problem of AI policy design, in Simeone, L., Gray, C. M., Verhoeven, A., de Götzen, A., Bakırlıoğlu, Y., Zohar, H., Stead, M., and Buwert, P. (eds.), DRS2026: Edinburgh, 8–12 June, Edinburgh, United Kingdom. https://doi.org/10.21606/drs.2026.1686
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Included in
The wicked problem of AI policy design
Development of policy tools and instruments has always trailed behind technological innovations. More recently, the rapid advancement of AI technologies has resulted in significant widening of this policy gap — a genuine wicked problem. To address this we developed a participatory AI Policy Design toolkit, engaging 23 technology researchers and practitioners to (1) critically evaluate existing AI use cases to identify key considerations for policy; (2) generate a catalogue of current and near future AI use-cases in their domains of expertise; and (3) collaboratively develop AI policy intervention proposal artefacts, and foster a shared vocabulary for AI policy design. Results highlight key tensions, challenges, and opportunities for collaboratively exploring AI policy discourse. The central contribution of this work is to help identify AI policy blind spots, equity concerns, and anticipate enforcement gaps, and building capacity to help democratize discourse about AI policy through contextually relevant, generative, and reflexive approaches.