Do You Have to Tell Customers You Use AI? From December, Yes

Transcript

This is Field Notes — your two minutes on the AI news that actually matters to Australian business. Today: Do You Have to Tell Customers You Use AI? From December, Yes.

Government data shows Australian SMEs check AI internally but rarely tell customers. From 10 December, that transparency gap becomes a Privacy Act liability.

Half of the Australian businesses using AI check its output before it reaches a customer. Far fewer tell the customer AI was involved at all. From 10 December 2026, that second number is the one that carries a penalty. Here are the four things to take away. First. The National AI Centre's SME AI Pulse puts Australian SME AI adoption at 43% (Dec 2025–Feb 2026) — but the safeguards that face customers lag far behind the ones that protect the business internally. Second. Roughly half of AI users check outputs before they reach a customer; telling customers AI is involved, and giving them a way to raise concerns, are much rarer. Third. From 10 December 2026, the Privacy Act requires disclosing automated decisions that significantly affect people in your privacy policy — and the OAIC began sweeping privacy policies in January 2026. Fourth. The Australian Consumer Law's misleading-conduct ban has no turnover threshold, so even Privacy-Act-exempt small businesses are liable if they overstate their AI or let a chatbot mislead.

So what does this actually mean for your business? That's the short version — enough to know whether it's worth a closer look this week.

That's your brief for today. Thanks for spending a couple of minutes with me — the links are in the show notes. Field Notes is brought to you by Coterie Labs. I'm Matilda. Until next time.