LEOPRD has published research on how often brands are excluded or overlooked in AI-generated shopping recommendations, suggesting this could affect online retail demand in Australia.
The report found that, in responses to unbranded category questions, the monitored brand was absent from 36% of answers and mentioned but not chosen in a further 35%. Only 29% of answers made the monitored brand the primary recommendation, based on LEOPRD's analysis of 31,200 AI responses from a broader monitoring programme covering 27 brands, seven categories and eight AI platforms.
Alongside that brand-monitoring work, LEOPRD analysed 420,903 anonymised prompts from Prompt Cowboy and found that 44% involved users asking AI to support information gathering, research, advice or judgement. A further 23% explicitly asked AI to compare, evaluate, validate or recommend a decision.
Australian users appeared more likely than those in the UK and US to ask AI to check or confirm a decision. The research found that 16% of Australian prompts sought that kind of validation, compared with 13% in the UK and 11% in the US.
LEOPRD has attached a commercial model to those findings. Its base case links brand exclusion from AI answers to A$2.9 million in potential exposure for every A$100 million of Australian online retail spending, with a sensitivity range of A$2.2 million to A$8.0 million. These are exposure estimates rather than measured revenue losses.
It also estimated that, at market scale, about A$2.4 billion in annual Australian online retail spending could be exposed where AI use, recommendation follow-through and brand exclusion intersect.
Decision support
The report argues that AI is moving beyond retrieving information and towards shaping consumer choices before a shopper reaches a retailer's website or sales channel. Requests for AI help with actions involving a vendor, service or commercial process doubled between March and August 2026, rising from about six to 12 per 1,000 prompts.
Those prompts included preparing outreach, creating vendor requirements, structuring interactions with providers and taking steps within a purchasing workflow. While that behaviour remains relatively limited, the report argues it shows AI moving closer to transactions.
"When AI answers a question about a brand and gets it wrong, someone pays. The customer can waste money and time. The business can wear the return, complaint or support call and when a physical product is bought, shipped and returned unnecessarily, there is a cost to the planet too with unnecessary packaging, transport and waste," said Celia Harding, Founder, LEOPRD.
The report says the commercial issue is not limited to factual errors. It argues that AI systems may also exclude a business if a user provides too little information, or if the public evidence available to the model is too thin or unconvincing to support a recommendation.
Consumer trust
LEOPRD also cited separate experimental research on how readily people defer to AI advice. According to the report, participants in that study followed AI advice 80% of the time when it was incorrect and 93% of the time when it was correct, although the tasks were experimental reasoning exercises rather than shopping decisions.
Professor Ben Newell, Director of the Institute for Climate Risk & Response and Professor of Behavioural Science at UNSW Sydney, commented on that pattern.
"People are asking AI to help frame what they need and then judge their options. The risk is that a confident answer can carry a veneer of authority, encouraging people to accept it without questioning the assumptions or evidence behind it. That goes beyond cognitive offloading. It becomes cognitive surrender," said Professor Ben Newell, Director of the Institute for Climate Risk & Response and Professor of Behavioural Science at UNSW Sydney.
Harding said the shift matters because AI does not need to make the final purchase decision to shape the shortlist consumers consider.
"AI doesn't have to make the final decision to influence it. If it tells someone which three companies are worth considering, validates the option they're already leaning towards or gives them a reason to rule another one out, it has already influenced the commercial outcome," said Harding.
Wider responsibility
The report argues that responsibility for how a brand appears in AI answers should not sit with marketing teams alone. It says the evidence available to AI models is produced across an organisation, including product information, complaints handling, pricing, media coverage, reviews and public commentary from executives.
That means improving how a business is represented in AI-generated recommendations may require operational changes as well as changes to content or search strategy.
"A marketing team can optimise content, but it can't optimise away a product problem, repeated customer complaints or an absence of independent evidence. Businesses need to start treating what AI says about them as a commercial outcome: understand where they're being excluded, why another option is being chosen, fix the underlying evidence and then connect changes back to demand, returns, complaints, service contacts and churn," said Harding.