Black Friday10% off any legal service

Commerce Commission AI Guidance NZ Business 2026

Do you need to change how your business uses AI now? On 18 November 2026, the Commerce Commission released AI guidance for New Zealand businesses, and the reassuring answer is that you don't need to stop using AI. You need to make sure your existing Fair Trading Act 1986 and Commerce Act 1986 checks still work. Separately, if your AI use involves personal information, you should also check your Privacy Act 2020 obligations.

The guidance is a timely reminder that AI doesn't get a special exemption from consumer protection or competition law. If an AI tool writes your ads, suggests prices, recommends products, answers customer questions or helps make decisions, your business is still responsible for how those tools are used.

For many small businesses, the main job now is practical: update your AI governance policy, tighten your marketing review process, check your website disclaimers, and make sure customer-facing AI tools don't overpromise.

What Has Changed With The Commerce Commission AI Guidance?

The Commerce Commission has issued new AI guidance explaining how existing consumer protection and competition laws apply when New Zealand businesses use AI. The key point is simple: the Fair Trading Act 1986 and Commerce Act 1986 still apply to AI-generated claims, AI-assisted pricing, AI recommendations, chatbots and other business tools.

Under section 9 of the Fair Trading Act 1986, you must not engage in misleading or deceptive conduct in trade. That means if your AI tool produces a product description, ad headline, chatbot answer or comparison table that gives customers the wrong impression, your business may still be responsible.

Under sections 10 and 11 of the Fair Trading Act 1986, you must not mislead the public about goods or services. Section 13 also prohibits false or misleading representations about matters such as price, quality, characteristics, uses, benefits and approvals.

The Commerce Commission's 18 November 2026 AI guidance also points businesses back to substantiation. Under section 12A of the Fair Trading Act 1986, you must have reasonable grounds for certain representations when you make them. So if your AI-generated product page says a supplement is "clinically proven", or your software page says your platform "cuts admin by 60%", you need evidence before that claim goes live.

AI can be useful, but it can also sound confident when it's wrong. Treat AI output like any other business communication: check it before customers see it, especially where it affects price, availability, product features, recommendations, competition-sensitive decisions or legal rights.

When Did The AI Guidance Start Applying?

The Commerce Commission released the AI guidance on 18 November 2026, and it applies as guidance from that date. The underlying legal duties aren't new because they come from existing laws, especially the Fair Trading Act 1986 and Commerce Act 1986.

That distinction matters. You don't need to wait for a new AI Act before reviewing your systems. If your business is already trading in New Zealand, your customer communications are already covered by the Fair Trading Act 1986 where they're made in trade, and your competitive conduct is already covered by the Commerce Act 1986.

Separately, the same practical urgency can apply to personal information. Under section 22 of the Privacy Act 2020, the information privacy principles apply to agencies that collect, hold, use or disclose personal information. If your AI tool collects customer details, summarises support tickets, analyses purchase behaviour or sends customer data to a third-party AI provider, the Privacy Act 2020 may be relevant now.

For example, Information Privacy Principle 1 in section 22 of the Privacy Act 2020 says personal information must be collected for a lawful purpose connected with your functions or activities, and the collection must be necessary for that purpose. Information Privacy Principle 3 requires you to take reasonable steps to tell people certain things when you collect information from them, including why you're collecting it and who will receive it.

So the practical deadline isn't a future commencement date. It's now. If AI is already inside your marketing, ecommerce, SaaS product, pricing process or customer service workflow, you should review it against the Commerce Commission's 2026 AI guidance and the existing legislation that sits behind it.

Who Does The Commerce Commission AI Guidance Affect?

The guidance affects your business if you use AI in a way that influences what customers see, pay, choose or understand, or how your business competes in the market. That includes obvious tools like chatbots, but it also includes quieter uses of AI in your back office.

You should pay attention if your business uses AI for:

  • writing product descriptions, ads, emails, landing pages or social media posts
  • generating pricing suggestions, discounts or personalised offers
  • recommending products, plans, add-ons or upgrades
  • answering customer questions through a chatbot or helpdesk tool
  • summarising customer complaints, reviews or support tickets
  • screening leads, bookings, applications or service requests
  • using third-party software to monitor competitor prices or adjust your own prices
  • building SaaS features that customers use to make business or consumer decisions

Under section 9 of the Fair Trading Act 1986, the focus is whether conduct in trade is misleading or deceptive, not whether a person or an AI tool wrote the content. That's why "the AI said it" won't be a clean answer if a customer is misled.

The Privacy Act 2020 can also come into play if personal information is involved. Under Information Privacy Principle 5 in section 22 of the Privacy Act 2020, you must protect personal information with reasonable security safeguards against loss, unauthorised access, use, modification or disclosure. If you paste identifiable customer data into an external AI tool without checking where that data goes or how it's stored, that can create a privacy risk.

This doesn't mean every AI use is high risk. Using AI to brainstorm internal blog topics is different from using AI to tell customers which medical appointment, financial plan or paid subscription suits them. The more your AI output affects customer decisions, prices or market conduct, the more careful your review process needs to be.

What Do You Need To Update Now?

You should update the documents and processes that sit around your AI tools, not just the tools themselves. A practical AI compliance check should cover the following steps.

  1. Map Where AI Is Used: List every AI tool used in marketing, sales, ecommerce, pricing, support, operations and product features, including free tools staff may be using informally.
  2. Classify The Risk: Mark each use as low, medium or high risk based on whether it affects price, availability, recommendations, customer eligibility, product claims, competitor information or personal information.
  3. Update Your AI Policy: Set rules for who can use AI, what information must not be entered, when human review is required, and who signs off public-facing content.
  4. Review Marketing Claims: Check AI-generated statements against sections 9, 12A and 13 of the Fair Trading Act 1986 before publication, especially claims about savings, performance, comparisons and endorsements.
  5. Check Pricing Controls: If AI suggests or changes prices, check how it uses competitor data, whether staff understand the inputs, and whether the tool could create Commerce Act 1986 risks.
  6. Check Website Disclaimers: Make sure disclaimers are clear, visible and consistent with what the page actually says, rather than trying to undo an overstatement buried elsewhere.
  7. Test Customer-Facing AI Tools: Run sample prompts through chatbots, product recommenders and helpdesk tools to check whether they give inaccurate, unfair or overconfident answers.
  8. Review Privacy Notices: If AI involves personal information, check your collection notices, privacy policy and information security policy against Information Privacy Principles 1, 3, 5, 10, 11 and 12 in section 22 of the Privacy Act 2020.
  9. Keep Evidence: Keep records showing what claims you checked, what evidence you relied on, and when you approved AI-generated content.

Your AI policy doesn't need to be long, but it does need to be clear enough for staff to use. A useful outline might include approved tools, prohibited inputs, human review, Fair Trading checks, competition checks, privacy checks, record keeping and escalation steps.

That outline isn't a finished legal document. Templates often miss the details that matter most for your business model, especially if you run an ecommerce store, SaaS product, regulated service or high-volume customer support team.

Worked Example: An AI Product Recommender In An Online Store

Say Mia runs an online skincare store in Auckland. On 19 November 2026, she turns on an AI product recommender that asks customers 6 questions and suggests a bundle. One recommendation says a serum is "guaranteed to clear acne in 14 days" and another says a bundle is "usually $180, today $99", even though the bundle has never been sold for $180.

Mia should pause those outputs before they go live. Under section 13 of the Fair Trading Act 1986, claims about performance characteristics, benefits and price can be false or misleading if they're not accurate. Under section 12A of the Fair Trading Act 1986, Mia should have reasonable grounds for objective claims such as a guaranteed result. A better process is to create approved claim wording, require a staff member to review the AI's recommendations each week, and keep evidence for any savings or performance claims.

What Are The Commerce Act Risks With AI?

The Commerce Commission's AI guidance isn't just about customer-facing advertising. It also flags competition-law risks under the Commerce Act 1986, especially where businesses use algorithms, automated pricing tools or shared third-party systems that influence market behaviour.

For small businesses, the main concern is usually not the technology itself. It's whether the tool helps create conduct that would be risky if humans did it. For example, cartel conduct can involve competitors fixing prices, restricting output, allocating customers or rigging bids. Using AI, software or a third-party pricing platform doesn't make that conduct safer.

Algorithmic pricing needs particular care. If your business uses a tool that monitors competitor prices and automatically changes your own prices, you should understand what data the tool uses, who else uses it, and whether it could lead to coordinated pricing outcomes. You should also be careful about sharing competitively sensitive information, such as future pricing, margins, customer lists, sales volumes or planned discounts, with competitors or platforms that competitors can also access.

This doesn't mean you can't use AI pricing tools. Many businesses use software to manage stock, demand, promotions and margins. But if a tool is designed to align pricing between competitors, relies on non-public competitor information, or removes meaningful independent decision-making, it's a risk area to flag before you rely on it.

Can You Rely On An AI Disclaimer In New Zealand?

You can use an AI disclaimer, but you shouldn't rely on it as your main protection. Under section 9 of the Fair Trading Act 1986, the question is whether your conduct in trade is misleading or deceptive overall, so a disclaimer may not fix a headline, chatbot answer or product claim that gives customers the wrong impression.

A useful AI disclaimer can still help set expectations. For example, your website might say that an AI chatbot provides general product information, may not cover every circumstance, and that customers should check key details such as price, availability, delivery timeframes and product suitability before ordering.

But the disclaimer needs to match the experience. If your chatbot gives very specific answers like "this plan is definitely best for you" or "delivery will arrive tomorrow", a small note elsewhere saying "AI responses may be inaccurate" may not be enough. Under sections 10, 11 and 13 of the Fair Trading Act 1986, the actual impression created by statements about goods, services, price and benefits matters.

For privacy, a disclaimer isn't the same as a proper collection notice. Under Information Privacy Principle 3 in section 22 of the Privacy Act 2020, when you collect personal information from someone, you generally need to take reasonable steps to tell them why it's being collected, who will receive it, whether collection is required or voluntary, and their rights of access and correction.

A good AI disclaimer should be:

  • clear: plain English, not technical wording customers won't read
  • visible: placed near the chatbot, recommender or AI-generated output
  • accurate: aligned with how the tool actually works
  • specific: explaining what the AI can and can't do
  • backed by controls: supported by human review, testing and escalation paths

The safest approach is to treat disclaimers as one layer of protection, not a substitute for accurate content and proper review.

Key Takeaways

  • The Commerce Commission released AI guidance for New Zealand businesses on 18 November 2026, confirming that existing Fair Trading Act 1986 and Commerce Act 1986 obligations still apply when you use AI.
  • Your business can still be responsible for misleading AI-generated ads, product descriptions, chatbot answers, recommendations and price claims.
  • Objective claims generated with AI still need reasonable grounds before you publish them.
  • AI pricing and decision-making tools can create Commerce Act 1986 risks, especially around algorithmic pricing, cartel conduct and sharing competitively sensitive information.
  • If AI tools collect, use, store or disclose personal information, separately check your processes against the information privacy principles in section 22 of the Privacy Act 2020.
  • The practical next step is to update your AI policy, marketing review process, website disclaimers, pricing controls and customer-facing AI testing records.

Official Sources

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

Get your customer-facing terms right

What should your privacy and online terms cover?

If you collect customer data, sell online or run marketing campaigns, your public terms and privacy documents should match the real customer journey.

Alex Solo
Alex SoloCo-Founder

Alex is a co-founder of Sprintlaw. Alex previously worked at a top-tier firm as a lawyer specialising in technology and media contracts, and founded a digital agency which he sold in 2015.

Get your customer-facing terms right

Get in touch with our team

Tell us what you need and we'll come back with a fixed-fee quote - no obligation, no surprises.

Need support?

Need help with your business legals?

Speak with Sprintlaw to get practical legal support and fixed-fee options tailored to your business.