The legal industry in New Zealand has always been more cautious than most about adopting new technology. That caution comes from genuine concern about privilege, confidentiality, and the duties lawyers owe their clients. But the picture is shifting quickly. Walk into most mid-tier Auckland or Wellington firms today and you will find AI being used somewhere in the back office, even if the partners are still careful about what they tell clients.

This matters to you as an NZ business owner for a simple reason. The firms you pay by the hour are about to become faster and more competitive. That is good news on your invoice. It also creates new risks you need to understand, particularly around privacy and how your data gets handled. The rules in New Zealand are not the same as Australia, the UK, or the US, and that difference matters.

What AI actually looks like inside a New Zealand law firm

When I sit down with firm partners around the country, the picture is fairly consistent. The headline-grabbing international tools get most of the attention, but most local firms are running a mix of general productivity software and one or two legal-specific products.

On the general side, the usual suspects are showing up in firm workflows. Microsoft Copilot is rolling out across firms that already sit on the Microsoft stack. Google Workspace customers are trialling Gemini. A small but growing number of firms, particularly in Christchurch and Tauranga, are using ChatGPT Team or Enterprise plans for internal research and first-draft work.

For legal-specific work, the tools we see most often include document review platforms that scan bundles of emails and contracts for relevant clauses, legal research assistants that pull case law faster than a junior can, and contract lifecycle tools that sit alongside the firm’s existing practice management software.

A partner at a mid-sized Auckland firm told me recently that roughly a third of their back-office time used to be spent on tasks they would now hand to a tool. The catch, and it is an important one, is that the partner still has to check the output. That is the part most clients never see on the invoice.

The use cases that actually affect your bill

Here is where the practical effect lands for you as someone paying the invoice.

Contract review is the most obvious. If you have ever sent through a 40-page commercial lease or a supplier agreement and waited a week for marked-up comments, expect that timeline to compress. Tools can surface unusual clauses, missing protections, and inconsistencies across documents in minutes rather than days. You should still expect a lawyer to read it, but the billable hours on that review should drop materially.

Due diligence on an acquisition is another. If you are buying a business, the legal team typically works through hundreds of documents, looking for change of control clauses, litigation mentions, and IP issues. AI cuts the initial pass down significantly. An advisor in our Christchurch network recently described halving the time on a particular deal, with part of the saving passed on to the buyer.

Legal research is the third. New Zealand case law is now well-indexed in several tools, and the time a junior lawyer spends pulling authorities has fallen sharply. For you, this means faster answers to specific questions, particularly on employment or commercial disputes.

What does this mean in dollar terms. For a typical SME legal matter, an hour saved on a task billed at $450 to $650 per hour, which is in the range we see for partners at mid-tier NZ firms, is real money. Industry estimates suggest AI-assisted legal work is somewhere around 20 to 40 percent faster on the tasks it covers well, but verify with your lawyer what savings they are actually passing through to you. The lower end of the saving is more common in our experience than the higher end.

The privacy rules you need to understand

This is the part that concerns me most, and it is where NZ law is genuinely different from the rest of the world.

The Privacy Act 2020 governs how any agency, including a law firm, handles personal information. There are 13 privacy principles. The one that bites hardest for AI is Privacy Principle 12, which deals with disclosure of personal information outside New Zealand.

If your law firm pastes a contract or an email chain into a public AI tool, and that tool sends the data to servers in the United States or another offshore location, that is an offshore disclosure under PP12. Your firm has obligations around that. They are meant to either get your consent, ensure the overseas recipient is subject to similar safeguards, or take other steps under PP12(2).

In practice, what this means for you is simple. Ask your firm what tools they are using, where the data is stored, and whether they have done a privacy impact assessment. If they are using consumer-grade AI and pasting your information into it without thinking, that is a problem you want to know about now rather than when the Privacy Commissioner writes to you.

A few other points worth knowing. The Privacy Act covers personal information, which includes names, emails, signatures, and contract terms involving identifiable people. It does not matter that the data is inside a contract rather than a database. The Act still applies.

If you are in the health sector, additional rules apply through the Health Information Privacy Code 2020. If you are working across the Tasman with Australian clients, suppliers, or staff, APRA’s CPS 234 covers information security for regulated entities, and your bank or insurer will have their own expectations of suppliers. For AU financial services work, ASIC’s RG 265 on social media and information handling can also become relevant, particularly around how AI-generated client communications are supervised. Verify with your advisor how these interact with your specific situation.

The point is not to scare you off AI. It is to make sure the firms and suppliers you work with have thought this through. The good ones have. Many have not.

The risks that sit outside privacy

Privacy is not the only concern. There are three others worth raising with your lawyer.

The first is accuracy. AI tools still get things wrong, particularly on NZ-specific law, which is a smaller jurisdiction and less represented in training data than the US or UK. A clause that an AI invents because it sounds plausible is a real risk. Your lawyer needs to verify every output before it goes near a court filing or a signed agreement. We have seen cases where an AI confidently cited a case that did not exist, and the human reviewer caught it only by chance.

The second is privilege. Legal professional privilege is the protection that lets you speak openly with your lawyer. There is some uncertainty in NZ law about whether work product that an AI produced without meaningful lawyer involvement still attracts privilege. The Law Society and the courts have not fully settled this. Treat AI as a research assistant, not a lawyer, and keep a human in the loop on anything privileged.

The third is supervision. The Lawyers and Conveyancers Act 2006 and the Rules of Conduct and Client Care set out what NZ lawyers owe their clients. A lawyer who hands a matter over to an AI without proper supervision is potentially in breach. Ask the firm who is responsible for the work, and what their review process looks like.

What to ask your law firm before any AI is used on your matter

If you want a short list of questions to put to your firm, here is what I would suggest asking.

One, what AI tools are you using on matters like mine, and which tasks are they used for. Two, where does the data go, and is it stored in New Zealand or offshore. Three, have you completed a privacy impact assessment under the Privacy Act 2020. Four, who reviews the AI output before it reaches me, and what does that review involve. Five, are you passing any cost savings through to me, and if so, how.

If your firm cannot answer these clearly, that is a signal worth paying attention to. The firms doing this well will answer without flinching.

Choosing your own tools versus relying on your firm

Some business owners I work with ask whether they should buy their own legal AI tools and bring them to their lawyer, or wait for the firm to use theirs. My view is generally to let the firm lead, for two reasons.

The first is the privacy and privilege issues already covered. If you are using a consumer tool at your end, you may be creating the very disclosure problems the firm is trying to avoid. The second is that your lawyer is the one with the professional duty. They are responsible for the advice. If a tool you supplied produces something wrong, the liability picture gets murky fast.

There is one exception. For non-privileged research, such as getting a quick steer on a contract clause before you ring your lawyer, a paid consumer tool used on a non-confidential basis can be a useful starting point. Just do not put anything sensitive into it, and treat the output as a starting point for a conversation, not advice. If you are a sole trader drafting a basic terms of trade document before sending it to your lawyer, that is a reasonable use. If you are reviewing an employment dispute with names attached, that is not.

What this means for your business over the next year

The short version is that legal services are about to get more competitive on price and faster on turnaround, particularly for the bread-and-butter work most NZ SMEs need. Contract reviews, employment documents, standard commercial work, and due diligence should all become faster and in many cases cheaper. If you are doing a property transaction through a firm that still uses REA Group data manually, expect that workflow to change. If you are using Xero or MYOB and your accountant is working with a law firm on a structure, expect the back-and-forth to compress. If you are advertising a role on Seek and