Australian consumer protection laws are failing to keep pace with AI-powered dynamic pricing tools that can charge shoppers different prices based on their browsing history, location and personal behaviour, a retail technology expert warned.
RMIT University lecturer Dr Aayushi Badhwar said pricing algorithms are already adjusting what Australians pay in real time, and that the legal framework governing the practice has not kept up.
"Consumer laws need to actively start considering technology and data, not just pricing in isolation," Dr Badhwar said. "We can't have loose regulations that don't account for the ethics of technology."
Unlike standard dynamic pricing — where prices shift for all shoppers based on demand — personalised pricing means two people can view the same product simultaneously and see different prices. Dr Badhwar described this as a legal "grey zone" that regulators have yet to clearly address.
The practice is already widespread, with Dr Badhwar claiming it is most prominent within the fashion industry. Retailers can track which products a shopper views, how long they browse, what they add to a cart and what they abandon — and use that data to adjust prices or offers in real time. Shoppers who leave an item in their cart may receive a discount email within hours; others may find the same item's price has risen.
"The system is trying to figure out one thing: what is the most you are willing to pay?" Dr Badhwar said.
For instance, if someone is identified as a lower-income or highly price-sensitive shopper, a few things can happen. They might see more discounts, but only just enough to get them to buy. They might be shown cheaper products and fewer premium options. Or they might be targeted more frequently until they “give in”.
She warned this dynamic could systematically disadvantage lower-income shoppers, who may not receive better deals but instead be targeted more precisely to extract the maximum they will spend. "It's less about giving the best possible price, and more about finding the threshold where that person will convert."
The lecturer noted that AI is now very good at figuring out if a customer is just browsing or if they are likely to buy, how many times they need to see something before purchasing, and whether they wait for sales or buy immediately.
These small patterns might seem minor for one person, but, according to Dr Badhwar, when this is done across millions of shoppers, even tiny improvements in targeting can significantly increase revenue for brands. The concern, however, is that this can quietly “squeeze” consumers.
“If the system knows your comfort zone, it can keep prices just below that point. So instead of getting a better deal, you’re consistently paying the maximum you’re willing to pay.”
There’s also a two-way effect happening now. Dr Badhwar said that with newer AI shopping tools, people might start setting their own price preferences, like saying they only want to buy something if it drops to a particular price range.
“That feels empowering, but at the same time, you’re also giving brands very clear signals about your limits. So instead of guessing, they now know your range. And because sales and discounts are already carefully planned into pricing strategies, it doesn’t necessarily mean you’re getting a true bargain. It just means you’re being offered a price that fits within your comfort zone.
“The technology is improving very quickly, but the rules around fairness and transparency are still catching up."
Dr Badhwar said the trend is visible in fashion retail. During the recent barrel leg jeans trend, high-demand items on fast fashion platforms like ASOS remained at full price even as other products were heavily discounted.
“Normally, on a platform that big, something is always on sale because there are so many brands and styles. But when you searched specifically for that trend, there were barely any discounts. So even though the platform overall had sales, the high-demand, trending items stayed at full price.
Australia lags behind the United States in deploying fully autonomous shopping AI, but Dr Badhwar said the gap is narrowing quickly. AI-generated product summaries already appear at the top of Australian search results, and advertising platforms are serving different versions of the same campaign to different users based on behavioural data. Within five years, she predicts consumers may delegate purchases to AI agents that track prices and negotiate with retailers automatically — while retailers run their own AI systems in response.
The Australian Competition and Consumer Commission’s Digital Platform Services Inquiry has examined some aspects of this space, but Dr Badhwar said disclosure requirements alone are insufficient. She cited the inquiry's own findings that most Australians do not understand the data implications of agreeing to website tracking policies.
"When you read 'by browsing this site, you agree to data collection,' what does that really mean? The language used in these policies is not easy to understand unless you've spent time learning about it," she said.
She also pointed to a regulatory gap where conduct that would be considered illegal if arranged directly between businesses may not fall under the same rules when achieved through algorithmic systems. "That gap is exactly what needs attention," she said.
When asked whether AI pricing tools could be designed to reduce consumption rather than increase it, Dr Badhwar said the technology exists but the incentive does not. "Most AI pricing systems are designed to do one thing really well: maximise revenue. Any positive impact on overconsumption is usually a side effect, not the intention.
“The reality is we’re operating in a system where consumption drives revenue. So, most tools naturally lean toward profit over reducing demand.”