When brands hide behind AI, Malaysians push back
Ellison Fernandez, Executive Creative Director, Dentsu Creative Malaysia
AI is no longer the bright-eyed new intern who drafts captions and draws cats with laser eyes.
In Malaysia, it now sits quietly behind some of the most consequential brand decisions on pricing, product recommendations, refunds, content moderation, and everyday service touchpoints.
“AI fatigue” is setting in, even as adoption remains uneven. Beyond tech skepticism, we are seeing signals of growing sensitivity to intent, shortcuts, and accountability, shaped by strong online discourse, regulatory attention, and cultural nuance.
But Malaysian consumers are not rejecting AI outright. As its influence grows, the biggest question for brands is whether people trust how it is being used.
Malaysia’s AI Context Raises the Bar for Brands
Malaysia’s approach to AI has been deliberate rather than reactive, with a clear emphasis on ethical, inclusive, and accountable deployment.
The establishment of the National AI Office in 2024 signalled a coordinated national effort to balance innovation with public trust, supported by governance and ethics guidelines that prioritise fairness, transparency, and human-centric design.
This matters because when responsible AI becomes a baseline at a national level, consumers naturally expect the same discipline from brands.
AI use is no longer judged only by efficiency or novelty, but by whether it aligns with Malaysian values around trust, responsibility and accountability.
The Trust Gap in AI-Powered Brand Experiences
Consumers in the B2C category, in particular, may not consciously think about algorithms or models, but they are quick to sense when something feels artificial or misaligned.
This often shows up when:
• Responses feel overly scripted or emotionally disconnected
• Personalisation feels too precise, poorly timed, or unexplained
• Automation replaces judgment in moments that require empathy
High-involvement categories make this especially clear. Telco brands, for example, operate at the intersection of data, daily usage, and long-term commitment.
When AI misfires here, it could turn up as closing complaints too quickly, pushing irrelevant offers during service disruptions, or responding without context. This is where frustration can escalate fast and publicly.
The same pattern plays out across B2C sectors, including banking, retail, platforms, and FMCG, where consumers want speed, convenience, and relevance, but not at the expense of transparency, empathy, fairness, or authenticity.
Consumers Push Back When AI Gets It Wrong
AI fails when it signals that efficiency matters more than care, or optimisation matters more than understanding.
We see this when automated service resolves issues efficiently but ignores emotional context; where algorithmic personalisation feels poorly timed or uncomfortably precise; or when it appears generic or tone-deaf during sensitive moments, such as service disruptions or national events.
In Malaysia’s culturally nuanced and highly social media-driven environment, these missteps travel fast.
The lesson applies particularly so for B2C brands, such as banks, retailers, telcos, platforms, and FMCG that are experimenting with AI-led engagement, with some examples of network disruptions in Malaysia witnessing firsthand how poorly handled interactions spread faster than apologies, with these experiences circulated widely online.
Where AI Needs Boundaries to Earn Malaysian Trust
AI gives brands in Malaysia an opportunity to move beyond personalisation as a marketing tactic and toward participation as a strategy to strengthen relevance, trust, and long-term loyalty.
• Design AI for real Malaysian needs
AI should help consumers understand trade-offs and make better decisions, not simply faster ones. This means reflecting local realities, such as price sensitivity, shared family usage, multi-line accounts, and seasonal behaviour.
In categories like telco, utilities, or subscriptions, AI can move beyond pushing “best” plans and instead help households understand trade-offs by comparing data sharing across family lines, highlighting bill volatility, or flagging when a cheaper option better fits actual usage.
• Turn community behaviour into brand value
Malaysian consumers trust peer opinions, social proof, and shared experiences. AI can be used responsibly to surface reviews, tips, and consumer insights that reflect real use cases, shifting the brand to be the best curator of its community.
In retail or e-commerce, this could mean highlighting reviews from similar household profiles, usage patterns, or locations, rather than generic five-star ratings.
In FMCG, AI can elevate real use cases of how families adapt products for different occasions, or how value packs are stretched across weeks.
• Make personalisation feel collaborative, not extractive
AI-driven personalisation should evolve based on feedback, choice, and consent. Allowing consumers to influence recommendations, pause personalisation, or adjust preferences signals respect.
In banking or retail, this can be as simple as allowing users to adjust recommendation signals, pause personalisation during certain periods, or explicitly state preferences rather than having them inferred.
• Humanise digital care, not just automate it
AI can streamline support, but brands should design clear handoffs to human assistance for emotional, complex, or high-stakes situations.
For example, during service disruptions or billing disputes, Malaysian consumers often want acknowledgment before resolution.
Yet, many encounter chatbots that close tickets automatically once a script is completed, or continue pushing add-ons while the core issue remains unresolved.
A more effective approach is designing AI to support human care by flagging frustration signals, escalating repeated complaints, and handing over full context to a human agent who can respond with empathy, language preference, and situational awareness.
• Use AI to reflect culture, not flatten it
Malaysia’s linguistic diversity, social norms, and cultural sensitivities require more than global templates. For AI to reflect culture rather than flatten it, Malaysian brands need to hard-code context into their systems.
This means setting clear cultural guardrails around when AI should promote, soften its tone, or stay silent altogether; training AI on locally approved language styles; defaulting to a customer’s demonstrated language preference without repeated prompts; and requiring human review for AI generated communications during high-risk moments such as service disruptions, crises, or sensitive national periods.
Just as importantly, brands should design escalation triggers that account for cultural and emotional context, and track signals like complaint recurrence, social backlash, and opt-outs to detect misalignment early.
When these controls are in place, AI stops sounding like a system optimised for efficiency and starts behaving like a brand that understands Malaysians.
The Real Test of AI Is Still Human
AI is now part of the background of everyday brand experiences in Malaysia. When it works well, it feels seamless. When it does not, it is immediately felt. As AI becomes infrastructure rather than innovation, trust will be the defining currency.
The brands that succeed will be those that balance automation with accountability, and efficiency with cultural understanding, using AI to support genuine connection rather than replace it.
(JUN/QOB)