Position: Prompts for Public-Sector LLMs Should Be Governed as Commons

Year: 2026

ICML · Peer-reviewed position paper

Author: Rashid Mushkani

Original bar chart comparing outcome proportions across prompting methods. — Figure 2 from the paper: outcome proportions by method in the pilot. Source

Mushkani, R. (2026). Position: Prompts for Public-Sector LLMs Should Be Governed as Commons. Proceedings of the 43rd International Conference on Machine Learning, PMLR 306, 171572–171582.

Prompt Commons treats the prompts used in public-sector AI as artifacts subject to public governance. The paper proposes versioned community repositories with records of origin, licensing, and moderation, and examines how stakeholder prompts can inform negotiated recommendations.

Read the proceedings entry.