What Is Generative AI? A Guide for New Zealand Businesses

Generative AI is a type of artificial intelligence that creates new content — text, images, code or summaries — in response to a prompt, rather than just classifying or predicting from data. Tools like ChatGPT, Microsoft Copilot and Claude are generative AI. For New Zealand businesses it can draft, summarise, answer questions and speed up knowledge work, but it needs governance around accuracy and data privacy to use safely.

How is generative AI different from traditional AI?

Traditional AI mostly analyses and predicts — spotting patterns, classifying data, forecasting. Generative AI produces something new: a paragraph, a summary, an image, a draft reply. The practical difference is that generative AI can do open-ended knowledge work a rules-based system never could, which is why it has spread so quickly through everyday business tasks.

Where does generative AI add value for NZ businesses?

• Drafting and rewriting — emails, documents, proposals and content.
• Summarising — long documents, meetings and correspondence.
• Answering questions — from your own policies, procedures and records.
• Speeding up analysis and first-pass research, with human review.
• Handling routine enquiries, escalating anything sensitive to a person.

What are the risks, and how do you manage them?

Two risks matter most. First, accuracy: generative AI can produce confident, wrong answers, so consequential output needs human review. Second, data privacy: entering personal or confidential information into public tools can breach your obligations under the New Zealand Privacy Act. Manage both with a clear AI-use policy, a sanctioned governed tool, and — for sensitive data — keeping processing onshore or inside your own infrastructure.

How should a New Zealand business start with generative AI?

Start narrow and governed rather than broad and open. Pick one or two high-value, low-risk tasks — drafting standard correspondence, summarising internal documents, answering questions from your own policies — and give staff a sanctioned tool plus a simple rule on what data may go into it. Keep a human reviewing anything that matters, and for sensitive or personal information keep processing onshore or inside your own infrastructure. Proving value on a couple of well-chosen tasks, safely, beats rolling generative AI out everywhere and hoping the risks look after themselves.

Will generative AI replace jobs in New Zealand?

The near-term reality for most New Zealand organisations is augmentation, not wholesale replacement. Generative AI takes over slices of work — first drafts, summaries, routine lookups — and hands the time back to people for judgement, relationships and the exceptions machines handle poorly. Roles shift rather than vanish: the valuable skill becomes directing and checking AI output well. Leaders who frame it as a tool that removes drudgery, paired with training on how to use it, get far better adoption and far less resistance than those who let staff assume it is aimed at their jobs.