Most Australians can’t tell the difference between AI-generated advertising and human-made creative, but two in five Australians say it would not impact their perception about a brand, according to new research by Ideally and commissioned by B&T.
The data suggests the advertising industry may be far more preoccupied with AI in advertising than the people actually seeing the ads.
The research, based on a nationally representative survey of 400 Australians, found only 12 per cent are “very confident” in their ability to tell the difference between advertising created by humans and advertising created using AI.
The remaining 88 per cent sit somewhere between uncertainty and total confusion — with 42 per cent saying they are “somewhat confident,” a third saying they are “not very confident” and 14 per cent being “not confident at all”.
AI ad impact on brand is negligible
Although consumers are mixed on their ability to tell if an ad is made by AI, the majority would like transparency.
Specifically, 27 per cent state it would ‘matter a lot’ and 40 per cent say it ‘would matter somewhat’. Just 13 per cent indicate it would not matter at all, highlighting the relevance of ad creation origin for most people.
Age and regional differences are notable. Youngsters (18-24-year-olds) are far more likely to say it matters “somewhat” (77 per cent), but are less concerned at the “a lot” level. Meanwhile older demos (those 55 and older) say it would matter “a lot”.
South Australians and Tasmanians are the most concerned, with 43 per cent indicating it would matter “a lot”, while responses in Victoria and WA/NT lean more towards “somewhat” and “not at all”.
Although most consumers do care about whether adverts were made with AI, it would not necessarily damage their perception of the brand.
Two-fifths of consumers reported that if they discovered an advertisement was AI-generated it would make no difference to their trust in the brand. Just under a third said they would trust the brand “slightly less”. Curiously, 8 per cent said they would trust the brand more.
Amaysim has doubled down on its AI-created adverts.
A quarter of women said they would trust brands “much less” if they discovered an ad was made with AI. Older consumers—31 per cent of 65-74 year olds and 35 per cent of consumers older than 75 years old—said they would also trust brands much less.
Those earning less than $75,000 are the most distrustful, with 31 per cent saying they would trust the brand much less if they knew an ad was AI-created, versus just 13 per cent of those earning $75,000-$150,000. Around the states, the Northern Territory and Western Australia show the least reduction in trust (49 per cent say no difference), while South Australia and Tasmania presented a slightly more sceptical profile.
As far as execution is concerned, again, most people couldn’t give a rat’s. A third don’t care and are entirely neutral.
A fifth feel “somewhat negative,” and 17 per cent feel “somewhat positive”. That said, 19 per cent said they would feel “very negative” and only a small minority of 8 per cent said they would feel “very positive”.
Younger adults (18-24) show higher positivity (32 per cent somewhat positive, 20 per cent very positive) compared to older groups, where negative views and neutrality increase with age—especially in the 55+ brackets.
Income also shapes attitudes: higher earners ($150,000) are more supportive (25 per cent somewhat positive, 17 per cent very positive), while lower earners lean negative. Gender splits are less pronounced, but men are more likely than women to view AI ads very positively (11 per cent vs. 5 per cent). Overall, while many are still undecided, positivity is more common among younger, higher-income groups.
In short, it seems all consumers want is to be told if their ads are AI. They don’t care either way, they’d just like to know.
‘The industry cares more than the consumer’
For agencies experimenting with generative tools, that gap between industry concern and public attention is already visible.
Tom Wenborn, chief tinker at Thinkerbell, says the debate inside the industry is far louder than the one happening outside it.
“We recently ran a test looking at selling a handbag with a real person vs an AI person. The real person won,” he said.
“For now it looks like there’s still a difference people are picking up on when it comes to people.”
Find out more: AI Adam Ferrier Flogs Bachelor’s Handbags (To Limited Success)
But beyond those cases, Wenborn says most audiences simply won’t dwell on the production process behind the work.
“However most people most of the time won’t care if AI has partially or wholly contributed to content,” he said.
“The industry cares more than the consumer, as it should, as we are in a state of transition and the industry needs to lean in and work it out.”
CEO and ECD of Paper Moose, Nick Hunter agrees.
Hunter told B&T that his take is “the customers don’t care unless they are told it’s AI,” and admits the technology has reached a point now where it’s “difficult to tell”.
He said for Paper Moose, AI has been “an incredible new tool”.
“It allows us to execute 800,000 shoots for a fraction so that is exciting. Challenger brands can compete on production values of big brands.”
The real risk isn’t the creative, it’s trust
While the debate has centred on AI-generated images, copy and creative execution, one governance expert says that’s largely the wrong focus.
Michael Schanker, chief marketing officer at OneTrust, says consumers rarely worry about whether AI created the ad itself.
“What consumers don’t care so much about is: was this ad copy or image generated by AI or not,” he told B&T.
“What they will care about is if they find that their data that they shared with the brand was used in a way that feels like a violation of their privacy.”
In other words, the creative execution is rarely the issue. According to Schanker, trust is.
“You spend as a company a long time building up trust and goodwill with customers,” Schanker said.
“But it only takes one thing to go very wrong.”
The bigger risk, he argues, lies in how AI tools are trained and what data feeds them — an area many companies and agencies are still navigating.
“Companies, agencies and even consumers don’t fully understand the risks yet,” he said.
“If you’re an agency you’re looking at AI tools that help you do more work faster, protect margins and still be creative. But you might not understand what that AI model was trained on.”
That lack of oversight, he warns, can quietly accumulate into serious problems.
“Companies might be sitting on ticking time bombs and have no idea,” he said.
“Because nobody’s really keeping an eye on the AI.”
Are the clients the real watchdogs?
Schanker believes if the consumers aren’t closely monitoring whether AI is used in advertising, the real pressure point may lie in agency – client relationships.
He said brands increasingly expect transparency around how work is produced — including “what tools are used and what risks may exist”.
“Agency work is based on trust with the company,” he said. “You always want to be transparent with your client about the work you’re doing, how you’re doing it, and what risks are involved.”
According to him, that means conversations about AI shouldn’t happen after a campaign is delivered, they should happen early in the process.
“I think most clients would want to know up front what tools are being used,” Schanker said.
Different industries will naturally have different levels of risk tolerance.
“A bank is going to be much more risk-averse than a food or beverage brand,” he said.
“So agencies should present options — maybe one route that uses more AI and another that uses less — and let the client decide the trade-offs in cost, speed and risk.”
As AI adoption accelerates across the industry, the research suggests the real conversation isn’t about whether audiences can spot AI-generated work.
It’s about whether brands, and their agencies, are being transparent about how it’s made.
