As AI becomes embedded in business operations: customer service chatbots, automated recommendations, AI-generated content: a practical question keeps coming up: do you have to tell customers when AI is involved? The answer depends on the context, but the general direction of regulation and expectation is toward more disclosure, not less.

What the law currently requires

Australia doesn’t yet have a specific AI disclosure law that applies broadly to small businesses. The Privacy Act 1988 requires organisations to be transparent about how they collect and use personal information: and if an AI system is processing customer data, that needs to be reflected in your privacy policy. But there’s no general legal requirement to tell customers “an AI wrote this” or “a chatbot answered your question.”

Sector-specific rules are different. The Australian financial services sector has obligations around automated decision-making. Healthcare providers using AI diagnostic tools face professional and regulatory disclosure obligations. Legal practitioners have duties of candour that may extend to AI use. If you’re in a regulated sector, check with your industry body or a lawyer about what disclosure is required.

What the ACL effectively requires

Even without a specific AI disclosure law, the Australian Consumer Law creates practical disclosure obligations. If a customer would reasonably expect to be dealing with a human: and is instead dealing with an AI: and that matters to their decision to engage with your business, not telling them could be misleading conduct. The ACCC has flagged this as an area of concern.

The clearest example: a customer contacts your “support team” expecting to talk to a person, gets an AI chatbot, and receives advice that turns out to be wrong. If they would have sought advice elsewhere had they known it was AI, there’s a potential ACL issue. This is particularly relevant for advice-giving contexts: health, finance, legal, insurance.

Where disclosure is clearly good practice

Even where it’s not legally required, disclosure builds trust. Customers are increasingly aware that AI is being used in business, and many prefer honesty about it. Specific situations where disclosure is clearly the right call:

AI chatbots handling customer service or support queries
AI-generated personalised recommendations (product, content, financial)
AI used to make decisions that significantly affect customers (loan approvals, insurance quotes, job applications)
AI-generated content presented as the business’s own expert opinion

What this means for your business

Update your privacy policy to reflect AI processing of customer data. If you use a customer-facing AI chatbot, label it clearly as AI. If AI is making or significantly influencing decisions that affect customers, consider telling them. You don’t need to disclose every use of AI: nobody expects you to announce that your email newsletter was drafted with AI assistance. But anywhere a customer might reasonably expect human judgment and is getting AI instead, disclosure is both good ethics and increasingly good risk management.

The regulatory direction globally is toward mandatory AI disclosure in more contexts. Getting ahead of it now is easier than retrofitting your processes later.

📬 Get the SmallBizAI.au Weekly

Practical AI news for Australian small businesses — every week, free.