Do You Need An AI Disclaimer In New Zealand?

Yes - if your New Zealand business uses AI-generated website content, tools or chatbots, you should usually have an AI output disclaimer.

A good disclaimer helps set expectations about what your AI can and can't do. It can explain that outputs may be incomplete, inaccurate or out of date, and that customers should check key details before relying on them.

But there's a catch. An AI disclaimer won't magically fix misleading website claims, poor customer terms or an AI chatbot that gives customers the wrong information. The real goal is to make your disclaimer work alongside your website terms and conditions, privacy wording and consumer law obligations.

What Is An AI Output Disclaimer?

An AI output disclaimer is a notice that explains the limits of content or responses generated by artificial intelligence on your website, app or customer platform.

You might need one if your business uses AI to generate or display:

• blog articles, product descriptions or guides
• chatbot responses to customer questions
• recommendations, comparisons or summaries
• automated quotes or eligibility answers
• health, finance, education, legal or technical information
• personalised suggestions based on customer inputs

The disclaimer should be written for the real customer journey. A tiny footer notice may not be enough if the AI output appears in a chatbot window, checkout flow or product recommendation page.

For example, if your chatbot suggests which skincare product suits a customer, the disclaimer should sit close to the chatbot or appear before the customer relies on the recommendation. If your AI-generated blog posts compare products, your disclaimer should be easy to find on those pages.

An AI disclaimer usually does three practical things:

• sets expectations by saying AI outputs can be wrong or incomplete
• limits reliance by telling customers to verify important information
• points customers to human help where the question needs professional or personalised advice

It shouldn't be used as a way to hide weak processes. If your business is publishing AI content, you still need review steps, clear website terms and sensible controls around what your AI is allowed to say.

Can An AI Disclaimer Protect You Under The Fair Trading Act?

It can help, but it won't protect you if the overall impression of your website is misleading or deceptive.

The Fair Trading Act 1986 applies to businesses in New Zealand when they promote, sell or supply goods and services. In simple terms, your business can't mislead customers or make false claims about things like price, quality, performance, availability, benefits or results.

This matters because AI tools can confidently produce statements that sound polished but aren't checked. If your AI-generated page says a product is approved, guaranteed, certified or suitable for a particular purpose, that statement needs to be true.

A disclaimer is only one part of the overall impression. If the main content makes a strong claim, a small disclaimer at the bottom of the page may not undo it.

Example: A Chatbot Makes The Wrong Promise

Say Mia runs an online skincare store in Wellington. Her AI chatbot tells customers that a $49 serum is safe to use during pregnancy and comes with a 30-day money-back guarantee.

In reality, Mia hasn't checked pregnancy safety with the supplier, and her website terms only offer returns for unopened products within 14 days. Over a weekend, 60 customers buy the serum after using the chatbot. That is $2,940 in sales linked to statements that may be inaccurate.

A line saying AI responses may not always be accurate probably won't solve the issue. Mia may need to correct the claim, update the chatbot, honour the stronger promise if customers reasonably relied on it, and review her website terms so they match what customers are being told.

The practical point is this: your AI disclaimer should reduce confusion, not contradict your sales pages or customer terms.

Does An AI Disclaimer Affect Consumer Guarantees?

Yes, but only in a limited way, because you generally can't contract out of core consumer rights for ordinary consumers.

The Consumer Guarantees Act 1993 gives consumers automatic guarantees for goods and services. For example, goods must be of acceptable quality and fit for purpose, and services must be carried out with reasonable care and skill.

If you sell to consumers, an AI disclaimer can't remove those guarantees. A disclaimer also can't excuse your business from providing services with reasonable care and skill just because an AI tool helped produce the output.

Where a disclaimer can help is by defining what your website content is and isn't. For example, you can make it clear that:

• AI-generated content is for general product information or navigation only
• the customer should check product labels, specifications and current pricing before buying
• the chatbot doesn't replace a consultation with a qualified professional
• final terms, pricing, delivery options and refund rights are set out in your website terms and conditions and checkout pages

If you sell business-to-business, there's more flexibility. Businesses can sometimes contract out of parts of the Consumer Guarantees Act where both parties are in trade, the exclusion is in writing, and it's fair and reasonable. This needs careful drafting in your customer contracts, especially if your website serves both consumers and business customers.

The safest approach is to avoid treating your AI disclaimer as a catch-all shield. Use it to explain limits, then make sure your actual customer experience meets the promises your website makes.

What Should A Chatbot Disclaimer Say?

A chatbot disclaimer should clearly tell users that the chatbot is automated, may make mistakes, and shouldn't be treated as final professional advice or a final statement of your business terms.

The exact wording depends on what your chatbot does. A simple customer service chatbot that helps users find shipping information may need lighter wording than a chatbot that gives health, financial, accounting, legal or technical guidance.

For many New Zealand business websites, your chatbot disclaimer should cover:

• automation: the user is interacting with an AI or automated tool, not a human team member
• accuracy limits: responses may be incomplete, inaccurate or out of date
• no professional advice: the chatbot doesn't replace advice from a suitably qualified professional
• verification: users should check important details, especially price, availability, suitability, warranties and delivery timeframes
• human escalation: users can contact your team if they need confirmation or personalised help
• privacy: users shouldn't enter sensitive or unnecessary personal information into the chatbot

You should also set rules behind the scenes. For example, your chatbot might be blocked from making medical claims, promising refunds, creating discount codes, giving legal or financial advice, or confirming stock levels unless it pulls from a reliable live system.

If your chatbot collects names, contact details, order numbers, health information or other personal information, you also need to think about the Privacy Act 2020. Your privacy policy should explain what you collect, why you collect it, who you share it with, where it may be stored, and how people can access or correct their information.

If your chatbot provider stores data overseas, you should also check your overseas disclosure obligations under the Privacy Act 2020 and what your provider's terms say about security, training data and retention.

How Should Your AI Disclaimer Fit With Website Terms And Privacy?

Your AI disclaimer should sit inside a wider set of website documents, rather than operating on its own.

Think of your AI output disclaimer as the warning label, not the whole safety system. It works best when it lines up with your website terms, privacy policy, product pages, customer service scripts and internal AI use rules.

For most businesses using AI on a website, the key documents are:

• Website terms: set the rules for using your site, including acceptable use, intellectual property, account rules, payment terms and limits on liability where allowed
• AI output disclaimer: explains the limits of AI-generated content, recommendations or chatbot responses
• Privacy policy: explains how you collect, use, store and disclose personal information under the Privacy Act 2020
• Customer terms: set the terms for buying your products or services, including pricing, delivery, refunds and support
• Internal AI policy: tells your team how to use AI tools safely, including review steps and prohibited uses

Make sure these documents don't contradict each other. If your AI disclaimer says chatbot answers aren't final, but your checkout page says chatbot offers are binding, customers may be confused and your risk goes up.

You should also place the disclaimer where people will actually see it. Good locations can include:

• next to the chatbot input box
• before a customer receives an AI-generated recommendation
• on pages that contain AI-generated articles or comparison content
• inside your website terms, with a short summary displayed near the AI tool
• during onboarding if customers use an AI-powered dashboard or app

If your business operates in a regulated or higher-risk area, such as health, financial services, accounting, education or legal information, get tailored advice before relying on a generic template. The wording needs to match your actual service model and compliance obligations.

AI Output Disclaimer Template: What To Include

A template can be a useful starting point, but your disclaimer should be tailored before you use it on your website. AI risks depend heavily on what your tool says, where it appears and how customers rely on it.

Here is a practical fill-in outline you can use when briefing your legal adviser or reviewing your current wording:

1. Identify the AI tool: state which parts of your website, app or service use AI-generated content or automated responses.
2. Explain the tool's purpose: describe whether the AI helps with navigation, customer support, recommendations, content summaries or another limited purpose.
3. State the limits of accuracy: say that outputs may be incomplete, inaccurate or out of date, especially where information changes quickly.
4. Limit reliance where appropriate: tell users to verify key details before making decisions, placing orders or relying on a recommendation.
5. Deal with professional advice: say the AI output doesn't replace advice from a qualified professional where the topic requires one.
6. Confirm priority of final terms: explain that your checkout page, customer agreement or written confirmation will apply over an AI response if there's an inconsistency.
7. Cover privacy basics: warn users not to enter unnecessary personal information and point them to your privacy policy.
8. Offer human support: explain how users can contact your team to confirm details or ask for help.
9. Keep records and updates: state that you may monitor, review and improve AI outputs, subject to your privacy obligations.

Common Mistakes To Avoid

The most common mistake is using a generic disclaimer that doesn't match the business. If your AI tool gives product recommendations, your wording should talk about recommendations. If it answers technical questions, your wording should address technical reliance.

Other mistakes to watch for include:

• hiding the disclaimer where users won't see it before relying on the AI output
• using broad wording that tries to exclude consumer rights that can't be excluded
• letting the chatbot make promises about refunds, warranties or delivery that your terms don't support
• forgetting to update your privacy policy when the chatbot collects personal information
• publishing AI-generated claims without a human review process
• treating the disclaimer as a substitute for checking your AI provider's terms

If something has already gone wrong, act quickly. Save the chatbot logs or page screenshots, pause the risky AI response, correct the public information, work out which customers may have relied on it, and get advice if the issue could involve misleading conduct, privacy concerns or regulated advice.

Key Takeaways

• An AI output disclaimer is usually a smart step if your New Zealand business uses AI content, recommendations or chatbots.
• A disclaimer can help set expectations, but it won't fix misleading claims under the Fair Trading Act 1986.
• You generally can't use an AI disclaimer to remove consumer rights under the Consumer Guarantees Act 1993.
• Your chatbot disclaimer should be visible, practical and matched to what the chatbot actually does.
• Your AI disclaimer should line up with your website terms, customer terms and privacy policy.
• Templates are only a starting point. The wording should be tailored to your business, your AI tool and your risk level.

If you'd like help with your AI output disclaimer, website terms, privacy wording or customer-facing AI processes, you can reach us at 0800 002 184 or team@sprintlaw.co.nz for a free, no-obligations chat.