Every trend deck this year has the same slide: consumers are turning against AI-generated marketing, Coca-Cola’s AI holiday ad got mocked as “soulless,” 78 per cent of people say AI makes ads feel less authentic, and the smart move for 2026 is to hide the AI and put humans back in front of the camera.
It’s a real, well-documented shift, mostly built on US and European data. And Southeast Asian marketers are quietly importing it wholesale, which is a mistake, because the region’s own data tells a meaningfully different story.
The narrative everyone is copying
The Western backlash is not exaggerated. Research unveiled at Cannes Lions in June 2026 by The Harris Poll, the 4As and Infillion found 78 per cent of consumers say AI makes ads feel less authentic, 73 per cent are less likely to trust an ad they suspect was AI-made, and 63 per cent are less likely to buy from a brand using AI-generated ads; more than two-thirds of consumers now view AI in advertising as largely a “marketing ploy.”
Coca-Cola’s AI-produced version of its “Holidays Are Coming” campaign drew enough public mockery that it became the go-to case study for what not to do. Brands are hiring people specifically to produce proof-of-human “behind the scenes” content, because “we made this ourselves” has become an actual selling point.
That’s a real pattern. It just isn’t Southeast Asia’s pattern.
What the region’s own data actually shows
Two things are true in Southeast Asia at once, and most trend pieces only report one of them.
First: the region is not reflexively hostile to AI production the way US audiences increasingly are, but the picture is more specific than “SEA doesn’t care.” A peer-reviewed 2026 study of 400 Gen Z social media users in Da Nang, Vietnam (Business Perspectives / Innovative Marketing journal) found that disclosing AI use actually increased trust (β = 0.469, p < 0.001), which in turn predicted purchase intention, a direct contrast to the Western pattern where suspected AI use erodes trust.
Also Read: The funnel was never neutral: What Asia’s markets reveal about Western marketing theory
A separate PLS-SEM study of 387 Indonesian TikTok users found AI labels raised viewers’ awareness of the ad but only reduced attitude toward hedonic, impulse-type products, not the broad rejection Western data shows across categories. The signal is consistent: SEA audiences aren’t punishing AI transparency, and in at least one documented case, honesty about AI actively builds trust rather than eroding it.
Second, and this is the part that actually matters commercially: whatever a brand’s stance on AI, human trust still overwhelmingly decides the sale. The 2026 eCommerce Influencer and Affiliate Marketing in Southeast Asia report from impact.com, Cube and Dentsu, based on 2,400 consumers across six SEA markets, found recommendations from family and friends are the single strongest purchase driver (2.42 out of 4), ahead of both online reviews (2.36) and creator recommendations (1.98), and that two in three respondents (67 per cent) had bought something specifically because a creator recommended it.
Generative AI has become a real part of product discovery, with 24 per cent of consumers now using tools like ChatGPT, Gemini or Claude to support shopping decisions, and adoption reaching 34 per cent in Vietnam and 31 per cent in Indonesia, the two fastest-adopting markets in the region. But online marketplaces still dominate both discovery (71 per cent) and where purchases are actually completed (88 per cent), and nearly half (49 per cent) of affiliate-driven purchases were motivated by trust and validation rather than price or convenience.
AI is a research layer sitting on top of a decision that trust, not algorithmic polish, still closes, and the report projects that by end-2027 roughly one in five SEA shoppers will complete a purchase natively through a generative AI platform, meaning this dynamic is intensifying, not settling.
Put those findings together and the regional picture looks meaningfully different from the Western one: SEA audiences aren’t reflexively punishing a brand for using AI, and disclosing it can even help. What they consistently punish is a brand with no human credibility behind it when the moment of decision arrives.
A related Cube study commissioned by Lazada found “authenticity-led” e-commerce, verified stores, trusted brands, credible reviews, has grown from 12 per cent of regional online retail sales in 2020 to 30 per cent in 2025, and is projected to reach 55 per cent by 2030. That is the real curve to watch, and it isn’t an anti-AI curve. It’s a pro-trust one.
The line isn’t AI vs human, it’s real vs performed
Where the region does draw a line, the Da Nang study offers a clue most trend pieces miss: the trust-building effect of AI disclosure was strongest among consumers who scored high on collectivistic orientation, meaning the value of admitting “this was AI-assisted” is tied to social and relational trust norms specific to the market, not a universal reaction to the technology itself.
That is a fundamentally different mechanism from the West’s “AI equals inauthentic” reflex. The offense in SEA isn’t the technology. Based on this research, it looks closer to deception: content engineered to disguise itself as an organic, personal opinion when it isn’t one.
This also tracks with a broader pattern in newer digital economies. Peer-reviewed research comparing consumer responses to AI-generated advertising in Vietnam and Australia, based on 839 collected responses, found the two markets process AI-made video ads differently enough that the study’s authors point to cultural dimensions like uncertainty avoidance as an underexplored factor.
Also Read: AI didn’t replace Southeast Asia’s marketing agencies, it repriced them
A market still in the process of establishing baseline digital trust in e-commerce evaluates new content differently than one that has already developed fatigue and suspicion toward synthetic media by default, which is closer to where the US now sits after roughly three years of AI-generated content flooding social feeds.
Several SEA markets simply haven’t reached that saturation point yet, and industry commentary from a recent Singapore ad-tech panel has floated the same idea from the operations side: that the region’s creator-anchored social commerce model may not follow the US trajectory into an “AI content flood” at all.
What this actually means for marketing teams here
- Stop copying the “hide the AI” playbook wholesale. The instinct to scrub every AI fingerprint from your output is solving a problem that current regional evidence suggests mostly exists in Western markets. Disclosing AI involvement has been shown to build trust in at least one SEA market, not erode it, so treat “we don’t hide our AI use” as a potential asset, not a liability.
- Protect the line that does matter: don’t fake spontaneity. AI-assisted product content that’s honest about being AI-assisted appears to perform fine, or better, with SEA audiences. AI-generated content staged to look like an unprompted, organic customer opinion is where the deception, not the technology, does the damage, and it’s also the area most exposed as platforms and regulators start scrutinising undisclosed synthetic endorsements.
- Treat creators and word-of-mouth as infrastructure, not a bolt-on. With family and friend recommendations and creator trust still outperforming both reviews and AI as purchase drivers, the ROI case for creator partnerships in SEA is arguably stronger post-AI than pre-AI, not weaker.
- Watch Vietnam and Indonesia specifically. They’re both leading regional AI adoption in product discovery and sitting in a digital-trust-building phase where credibility cues matter more than production origin. That combination rewards brands that pair AI-assisted efficiency with genuinely credible, verifiable claims, not brands that either over-rely on AI or over-correct into performative “100 per cent human” theatre.
The uncomfortable part for a lot of regional marketing teams
The convenient story is that Southeast Asia is simply a few quarters behind the US on the same backlash curve, and the smart move is to pre-empt it. The regional data doesn’t clearly support that. It suggests a market running on a different trust mechanism entirely, one where the scandal isn’t “a machine helped make this,” it’s “you tried to make a machine’s output pass as somebody’s honest opinion.”
Brands spending 2026 anxiously stripping AI fingerprints out of product content may be solving last year’s American problem while missing the one actually sitting in their own market: whether anything a real customer would vouch for is still standing behind the campaign at all.
—
Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.
The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.
Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.
The post The AI marketing backlash story doesn’t actually fit Southeast Asia appeared first on e27.