evidoria

← Back to browse

Good practice

NZCER's Generative AI in Aotearoa New Zealand Primary Schools — National Teacher and Student Survey

New Zealand · Wellington · See the New Zealand profile

NZCER's 2025 survey of 266 NZ primary teachers and 147 students found generative AI already widespread — 69% of teachers use it weekly — but poorly governed: 96% rely on free tools, only 44% feel confident teaching responsible use, and most students recall no adult guidance.

69 %
Teachers using generative AI weekly (2025 survey)
96 %
Teachers relying on free consumer AI tools (2025 survey)
44 %
Teachers confident teaching responsible AI use (2025 survey)
85 %
Teachers wanting more professional-learning support (2025 survey)
~90 %
Students who had heard of AI (2025 survey)
266 teachers
Teachers surveyed (2025)
147 students
Students surveyed (2025)
NZCER's Generative AI in Aotearoa New Zealand Primary Schools — National Teacher and Student Survey

Details

Promoter
New Zealand Council for Educational Research (NZCER)
Period
2025 (survey conducted and report published)
Keywords
education research, AI literacy, teacher professional development, primary education

Context

NZCER, an independent national education research body based in Wellington, published 'Generative Artificial Intelligence in Aotearoa New Zealand Primary Schools: Teacher and Student Survey Findings' in 2025 (Coblentz, Dong and Gibbs; DOI 10.18296/rep.0077). The study surveyed 266 primary-school teachers (Years 5–8) nationally and 147 students across four case-study schools, examining how generative-AI tools such as ChatGPT and Google Gemini are already being used in classrooms in the absence of a national policy framework.

Results

69% of teachers reported weekly generative-AI use for lesson planning, assessment and personalising materials; 96% relied on free consumer tools rather than school-funded premium licences; only 44% felt confident teaching students responsible AI use; and 85% wanted more professional-learning support. Among students, roughly 90% had heard of AI and over half were already using generative tools, but most could not recall being given any adult guidance on when or how to use it.

Conclusions

The teacher sample was self-selected and 'disproportionately interested in generative AI,' and the 147-student sample across four schools is a small, non-representative case-study set — so the findings describe an engaged, early-adopter slice of the system rather than a nationally representative baseline. NZCER recommends centrally funded professional learning, privacy-protected access to premium models, culturally grounded national and school policies, and staged, age-appropriate AI critical-literacy teaching — an explicit acknowledgement that New Zealand still lacks a coherent national AI-literacy policy for schools.

Implementation

Indicative cost
Low (< €50k) — This is a diagnostic research report rather than a deployed programme; cost_band set to low reflecting the survey/report's own scope.
Time to results
Short (< 1 year) — A single-year survey conducted and published in 2025; timeline_band set to short.
Staffing & skills
NZCER research team (Coblentz, Dong, Gibbs)

Conditions for success

  • NZCER's own recommendations: centrally funded professional learning, privacy-protected access to premium AI models, culturally grounded national/school policy, staged age-appropriate AI critical-literacy teaching

Common failure modes

  • Self-selected teacher sample 'disproportionately interested in generative AI'
  • Small, non-representative student sample (147 students, 4 schools)
  • No national AI-literacy policy currently in place

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful