AI for New Zealand school administration.

Where AI helps a school office

AI helps schools most where office staff and senior leaders handle a steady stream of predictable work: parent and whānau emails, absence notes, uniform and fee questions, enrolment packs, trip consent forms, newsletters, board papers and staff asking what the policy says. It does not belong in decisions about a child’s learning, wellbeing, behaviour or placement. This page covers school administration only.

Leaders are already using it, mostly informally. The Education Review Office’s 2026 report Ready or not, based on almost 4,000 survey responses from students, teachers, leaders and parents in Term 1 2026 across English-medium state and state-integrated schools, found 93% of school leaders using AI, most commonly for drafting and communications (80%). It also found around three in four schools in an “unplanned” category, about seven in ten leaders without a data privacy and security policy for teachers’ AI use, and 56% of leaders saying they do not know enough to use AI tools properly.

School administration workflow map

Worked guides cover the first three: triaging parent enquiries, building a staff policy assistant and administrative document workflows.

What the Ministry, NZQA and ERO say

• Ministry of Education. Its generative AI guidance, last updated 22 May 2026, tells schools not to use personal data in AI tools, because “what you put in may be reused.” It says AI tools for marking should support, not replace, teachers’ professional judgement, and that generative AI is not permitted in NCEA external assessments.
• NZQA. Schools with consent to assess DASS-listed standards must have an authenticity policy that includes the acceptable use of AI (page updated 18 September 2026). An office AI policy should sit alongside it, not contradict it.
• ERO. The 2026 report calls for a more deliberate, system-wide approach and shows most schools have not yet written guidance for staff.
• Privacy Act 2020. Information privacy principle 4 requires fair collection and gives particular weight to collection from children and young persons. The Privacy Commissioner expects a privacy impact assessment before a school adopts an AI tool, and wants a person to review outputs before anyone acts on them.

Given ERO’s finding that most schools lack a privacy policy for staff AI use, a one-page office policy is worth agreeing before any build. It should cover at least:

• Which AI tools and accounts staff may use for school work, and who approves new ones
• What must never go in: student names with health, behaviour, learning support or family information
• Who reviews AI drafts before they reach whānau, and which messages are never drafted by AI
• Where AI outputs that become school records are stored
• How the office policy fits the school’s NZQA authenticity policy

This is general information. Boards should take advice from their privacy officer, the Ministry or their school’s advisers on specific uses.

The systems schools run

Most schools use a student management system such as Kamar, Hero, eTAP or Edge, a policy platform such as SchoolDocs, and Google Workspace for Education or Microsoft 365. The safest early projects work from email and documents rather than pulling records out of the student management system, which holds the most sensitive data. Check the AI settings in your education accounts, who can switch features on, and whether inputs are used for training. Our notes on Google Workspace and Microsoft 365 cover the admin controls.

A first project: parent enquiry triage

The office inbox is high volume, mostly routine, and easy to measure, and staff keep the send button.

1. Write the approved answers. Collect term dates, fees, uniform, transport, absence and enrolment information into one maintained document set.
2. Define the escalation list. Any mention of bullying, injury, custody, health, learning concerns or distress goes straight to a named staff member, with no drafted reply.
3. Draft, never send. The system labels each message and drafts replies to routine ones from the approved answers; office staff edit and send.
4. Check contact restrictions. Messages about a student are checked against any contact restrictions recorded by the school before anything goes out.
5. Review monthly. Track response times, how many replies needed heavy edits, and any concern that was not escalated.

Illustrative example: an office receives about 80 enquiries a day and spends around 3 minutes on each, 240 minutes. If 60% are routine and staff approve drafted replies to those in about a minute each, the day’s work becomes 48 + 96 = 144 minutes, saving roughly 8 hours a week. The measure that matters most is zero missed escalations. Our AI workflow automation service builds this, and AI training helps office teams and leaders use it well.

Where AI fails in school administration

Workflows we map in this sector

• Staff policy assistantA private assistant that answers staff questions from your school’s own policies and agreements, with the clause cited.
• Parent enquiry triageSort the school office inbox by category and urgency, with learner details removed before any AI model sees a message.
• Administrative document workflowsExtract, check and file enrolment packs and other school forms, with staff confirming every record and nothing deleted.

Written by the Levy Street team in Wellington. Last updated 19 October 2026.