Half of Australians Say AI Made Customer Service Worse

Half of Australians say AI made their service worse

More than half of Australians say the rush to put AI into customer service has made it worse. Fewer than one in ten say it has improved the products and services they buy. Those figures come from consumer group CHOICE, which put the question to more than 1,000 Australians this month — and for any business that has just dropped a chatbot on its website or an automated voice on its phone line, they land as a warning. Your customers can tell. And in a trade or a professional practice, where the next job comes from the last one, a customer who felt handled by a machine is a referral you'll never hear about.

CHOICE's senior campaigns and policy advisor, Jordan Cornelius, called it a shift in power: AI, she said, "could result in one of the biggest shifts in power away from consumers we've seen", eroding their ability to assess value or shop around through surveillance pricing, dark patterns and "poor customer service experience". Strip out the policy language and the message to a small business is simpler. Australians aren't rejecting AI. They're rejecting being walled off from a person by it.

The survey is a map, not a verdict

Read the wrong way, that finding is a reason to keep AI out of the front of your business. Read right, it's a map. It tells you exactly where AI helps and where it does damage. AI earns its place when it captures contact you'd otherwise lose — the call that comes in while you're on the tools, the booking someone wants to make at 9pm, the same question you answer twenty times a week. It does damage the moment it stands between a customer and the thing they called about: the phone menu with no way out, the chat window that cheerfully cannot help, the "talk to a human" that goes nowhere.

The same AI phone agent that stops a tradie's call going to voicemail and off to a competitor turns into a liability the instant it becomes the only option and can't solve the problem. This is a Customer Intelligence question before it's a technology one: not "can we automate this?" but "what does this customer need from us right now, and can a machine actually give it to them?" Get that wrong at the front desk and no amount of saved admin time makes up for the customer who doesn't come back.

Where to automate, and where to keep a human

Draw the line by the kind of contact, not the channel. Hand to AI the work that's routine and high-volume: answering and booking overflow and after-hours calls, sending confirmations and reminders, chasing a document, giving a job status, answering the questions whose answers never change. It's the same front-desk logic that lets an allied health clinic fill the chair a no-show left empty — capture, not deflection. Keep a human, or a fast and obvious route to one, on everything that carries judgement or feeling: a complaint, a bill in dispute, a customer who's upset, a quote that needs a decision, anyone who matters enough that losing them would hurt.

The single thing that separates the businesses customers reward from the ones they quietly leave is the escape hatch. Every automated channel needs a visible, one-step way to reach a person — and it has to actually work. If you're not sure which of your own customer touchpoints are helping and which are hiding from the customer, map where a bot earns its place before you switch one on.

What to do before you switch anything on

Four steps, none of them technical. First, decide the one job the AI is for, and make it capture, not deflection — the goal is to catch the contact you're missing, not to stop customers reaching you. Second, write the handoff rule: at what point does a person take over, and how fast. Third, test it as a customer — ring your own line, open your own chat, and count the steps to a human; if you run out of patience, so will they. Fourth, measure the right thing. Not calls deflected or tickets the bot closed on its own, but jobs booked and complaints resolved. The businesses that win with AI in customer service aren't the ones that automate the most. They're the ones that automated the routine and kept a person on the moments that decide whether a customer comes back.