AI Policy Sandbox (AIPS)

About AIPS

New Zealand’s AI policy conversation rests on a fragmented evidence base. Adoption figures in circulation range from 32% to 87%. No two sources measure the same thing. Policy built on any single number is standing on weak ground.

AIPS replaces single-number thinking with a structured, sector-calibrated framework that compares different policy designs under uncertainty.

What it does

The sandbox compares three policy approaches across New Zealand’s full economy:

• Aggregate policy - broad economy-wide allocation by GDP share
• Targeted demand-side - support focused on sectors where adoption is lagging or bottlenecks are acute
• Targeted supply-side - investment in enabling capacity: technology, skills, infrastructure, and diffusion

Sector coverage

The model uses a tiered structure covering all 19 ANZSIC Level 1 sectors. Nine Tier 1 sectors carry full explanatory depth, each representing a distinct adoption archetype - from agriculture (no published adoption rate, long diffusion timeline) to technology (supply-side enabler, $24B GDP).

Current status

Model equations v0.1, v0.2, and v0.3 are all merged. All 19 ANZSIC Level 1 sectors now carry calibrated numerical parameters. The scenario specification and a YAML parameter registry schema are drafted and in review (PR #8). Next: first v0 simulation against the 9-run policy comparison set, then registry migration from CSV to YAML.

The target output is a peer-reviewed paper and an open-access interactive web tool.

Get involved

The project welcomes sector specialists, economists, policy analysts, and AI practitioners. See the full project page or explore the GitHub repository.