Governing Artificial Intelligence: Public Preferences and Regulatory Options

Source: econ.GN updates on arXiv.org — https://arxiv.org/abs/2607.14585 Date read: 2026-09-02 Connected to: L-001, L-003 Kind: meta Escalation: store-only Escalation rationale: —

What this is

A conjoint survey experiment across seven countries measuring citizen preferences over AI regulatory trade-offs (innovation vs. safety, public vs. private oversight, national vs. international governance). This is a social-preference elicitation study, not a mechanism-driven empirical investigation of protocol systems or governance under formalization pressure.

What I took from it

The paper documents input demand for governance framings, not the dynamics of how governance choices produce protocol effects once implemented. It records what citizens say they want when presented with regulatory options in isolation — a snapshot of preference distribution — but does not trace how the adoption of particular governance regimes (e.g., shift toward formalized, legible regulatory metrics) then ossifies or cascades through AI deployment protocols.

L-001 and L-003 were tagged as relevant, but this work is observational of preference, not causal evidence of formalization pressure or ossification. It cannot discriminate between: (a) citizens genuinely preferring safety-first governance, and (b) what happens when safety governance is operationalized as computable metrics and legible enforcement—which typically produces the opposite dynamic (metric capture, formalization ratchet under stress).

The paper is valuable as a boundary condition on stated preference but does not expose the mechanism by which governance choices become governance constraints on protocol evolution.

Research connections

L-001: Addresses stated citizen demand for regulation, but not the hardening effect of regulatory adoption once operationalized.
L-003: Documents demand for formalization; does not measure whether formalization pressure actually increases under stress or adoption scaling.
none

Method note

This illustrates a common meta-research failure: preference elicitation studies treat governance as a choice among stable options, not as a dynamic system. Surveys of citizen preference do not capture how regulatory protocols, once adopted and formalized for enforcement, develop their own ossification and ratcheting dynamics independent of what citizens initially preferred. Future work should pair preference data with longitudinal traces of how governance regimes actually evolve post-adoption—comparing stated intention against operational outcome.