AI in CX: Why Consumer Trust is Declining and How Brands Can Fix It

Only 5% of shoppers in APAC fully trust AI-generated brand content, says Klaviyo’s Germaine Tay. “The first warning sign is usually a drop in engagement quality,” she explains why AI overuse is eroding trust, and what brands need to change now.

There was a time when automation in marketing was seen as progress, a way to scale communication, improve efficiency, and reach customers faster than ever before. That equation is now starting to break.

Today, AI is everywhere. It shapes how brands communicate, how customers discover products, and how decisions are made. But as adoption increases, trust is moving in the opposite direction.

Consumers, particularly in APAC, are not just using AI more; they’re becoming more critical of it. They can recognise when automation adds value, and more importantly, when it doesn’t.

That shift is creating a new kind of tension for brands. The pressure to scale hasn’t gone away, but the tolerance for low-quality, automated interactions has dropped sharply. What was once seen as efficiency is now, in many cases, being perceived as noise.

“The breakdown comes down to brands prioritising efficiency over genuine value. Only 5% of shoppers in the region fully trust AI-generated brand content,” said Germaine Tay, Head of Marketing, Asia at Klaviyo.

In this conversation, Germaine Tay explains why trust in AI-driven marketing is declining, how “AI slop” is impacting brand perception, and what organisations need to change to rebuild meaningful customer relationships.

Full interview;

Your research points to a clear contradiction: consumers are using AI more than ever, but trusting it less. What, in your view, has fundamentally broken in that relationship?

The breakdown comes down to brands prioritising efficiency over genuine value. APAC consumers now use AI more frequently than their counterparts in the US and Europe, with 30% using it several times a week. This means they’ve become the world’s most discerning AI audience. They can spot automation that adds nothing to their experience, and they’ve grown tired of it.

Only 5% of shoppers in the region fully trust AI-generated brand content, compared to 12% in the US and 16% in Europe. That gap is a strategy problem. Brands that deployed AI purely to scale output without investing in quality or relevance have eroded the trust they worked years to build.

We’re seeing a rise in what you call “AI slop.” Explain this, and its possible impact for brands.

AI slop is low-quality, mass-produced automated content that feels hollow, generic, and disconnected from any real understanding of the customer. More than half of APAC consumers (51%) now regularly spot it in their social feeds and brand interactions.

For brands, the impact is stark, as it actively damages equity particularly among younger, digitally native audiences who have the highest expectations for authenticity. What makes this especially urgent is that the damage isn’t passive. Consumers aren’t just ignoring this content; they’re losing trust in the brands producing it.

In a market as competitive as APAC, where customer relationships are hard-won, that erosion is difficult to reverse.

What advice would you give a business leader to build marketing trust in AI?

Move from scaling output to scaling usefulness. Brands that use AI to solve real customer problems instead of simply filling more channels with more content are the ones that will continue to win their customers over.

That means investing in the data infrastructure that allows AI to actually be personalised, and not just in name. It also means keeping humans meaningfully in the loop. Singapore’s 2026 budget, which committed over S$1 billion to AI infrastructure alongside the establishment of a new National AI Council, reflects exactly this tension at a government level.

Having human-centric oversight enables sustained AI adoption rather than hinder it, and brands should take that same approach.

If customers can now easily spot low-quality AI content, what signals make these interactions feel inauthentic?

The most common signals are generic, repetitive content that lacks local or personal context, and interactions that feel scripted rather than responsive to who the customer actually is. APAC consumers have developed a finely tuned radar for content that could have been written for anyone because increasingly, it has been.

What’s particularly striking in the data is that this scepticism has created an identity crisis: 63% of shoppers have mistaken high-quality human-written content for AI, simply because they’re so primed to distrust brand messaging.

That’s a real problem for brands investing in authentic storytelling, because their best work is being discounted before it’s even properly read.

As AI increasingly influences how people discover and evaluate products, brands no longer have full control over the journey. How does that shift the role of marketing and CX teams?

The role shifts from controlling the journey to powering the intelligence behind it. With 78% of APAC shoppers already using AI to compare brands or get product recommendations, and 66% doing so specifically in electronics marketing, CX teams need to carefully consider the quality of the data and content they’re feeding into these systems.

If the underlying data is fragmented or inaccurate, the AI recommendations built on top of it will be too. This is where tight platform integration matters. When commerce data, customer behaviour, and marketing execution all operate from a unified foundation, the intelligence layer has what it needs to make recommendations that are actually useful to the customer, and not just what is algorithmically convenient.

From your perspective, what are the early signs that a brand’s use of AI is starting to damage, rather than build, long-term customer relationships?

The first warning sign is usually a drop in engagement quality. This doesn’t just mean open rates or clicks, but the kind of interactions that signal customers actually value what you’re sending them.

The second is when feedback, whether through support channels, social mentions, or direct responses, starts reflecting a sense that the brand feels less human than it used to.

But honestly, don’t wait for the signals. With behaviours shifting this fast, the most useful thing a leader can do is talk to their customers directly. A handful of honest conversations will tell you more than a dashboard will.

When 51% of your audience regularly identifies low-quality automated content in their feeds, the margin for error is thin, and the brands that catch it early are usually the ones where someone picked up the phone before the metrics moved.

How do we prepare for a future in which not only CX teams but also AI agents could design customer journeys? What does the infrastructure need to look like for this to be possible?

The foundation has to be a unified, real-time data infrastructure that keeps commerce and customer data connected at every touchpoint. Without that, AI agents are working with incomplete information, and the experiences they design will reflect that incompleteness.

What we’re seeing with Klaviyo and Shopify’s deeper integration is a practical example of this in action, connecting localised catalogue data, regional pricing, and market-specific content so that every layer of the experience draws from a single, accurate source.

Brands using both platforms saw 73% revenue growth over three years, reflecting what’s possible when the data foundation is solid.

Beyond the technical architecture, human-centric oversight will remain essential. AI agents can optimise and execute at scale, but the judgment calls around brand values, tone, and customer experience still require human direction.

That balance between strong infrastructure and thoughtful governance will define which brands use AI effectively and which simply use more of it.