Deploying AI or Being Deployed by AI ? Learning from Singapore

What a 700-square-kilometre nation can teach a mid-cap CMO about AI, sovereignty, and the cost of missing an industrial revolution.

I was witness last week, albeit digitally, when two of the sharpest minds in Asia-Pacific foreign policy sat down to talk about where AI sits in the arc of industrial revolutions. Singapore's Foreign Minister, Dr Vivian Balakrishnan, and Kevin Rudd — former Australian Prime Minister and President of Asia Society — were on the sidelines of the United Nations General Assembly in New York. The conversation covered geopolitics, strategic competition between the United States and China, demographics, and what it all means for smaller nations caught between the two.

They were talking about economies. My mind was drawing parallels to brands and companies.

Because the structural position of a small Southeast Asian economy in the age of AI is, I would argue, identical to the structural position of a small or mid-cap company in the global marketing battlefield. The Fortune 500 companies — the Unilever’s, the Coca-Cola’s, the Nestlé’s — are the superpowers. They have the budgets, the human capital, the data infrastructure, and the institutional memory to build ( or dictate ) pretty much whatever they need. The Russell 2000 companies, and their equivalents on every other bourse, are the Singapore’s and the Malaysia’s: high ambition, small population and resources, and the next industrial revolution at the door that could either accelerate their economies or leave them further behind the goliaths in their geo-political spheres.

Vivian described the moment this way: "For those of us who have made it into that competitive lead, or at least are part of that global supply chain, there remain major opportunities in the midst of major challenges." The qualification in that sentence — for those of us who have made it — is the entire question.

( Balakrishnan–Rudd conversation, Asia Society, New York, 25 September 2026. Published on Kevin Rudd's Substack )

I: Deploy, don't develop

The single most important line Vivian delivered — the one that maps most directly to the mid-cap marketing problem — was his distinction between developing and deploying AI. Singapore, he said plainly, does not believe Southeast Asia can compete with America and China in developing frontier models. The special sauce of Silicon Valley, New York, Beijing or Shenzhen exceeds what the region can generate. But Singapore can be at the forefront of deploying these tools.

This is not a concession of weakness. It is a strategic choice about where to compete — and it is the choice every mid-cap CMO must now make.

A Unilever or Coca-Cola will build proprietary AI stacks with exclusive partnerships with frontier models. It will invest hundreds of millions into custom models for media mix modeling, creative optimization, supply-chain intelligence, and demand forecasting. A company with a $50 million marketing budget will never do that, just as Singapore will never build a frontier foundation model to rival GPT or DeepSeek. And it does not need to.

What it needs is the architecture to deploy what already exists — and the capability to govern the deployment. That is an entirely different problem, and it is one where agility, not scale, is the advantage.

Singapore's GDP per capita rose from US$500 to over US$80,000 in 61 years — not by competing with China on scale, but by owning the critical capabilities and plugging into the global system from a position of selective sovereignty.

II: Sovereignty and the colonization risk

Vivian's career has been built around the sovereignty proposition: how a small nation retains strategic autonomy when the powers on either side of it could, in theory, dictate terms. His formulation, offered with the directness of a surgeon, was pointed: "If you're not at the table, you're on the menu. That's why we were colonized during the last industrial revolution."

Asia missed the last industrial revolution, and the consequence was centuries of foreign rule. China, he said, has absolutely no intention of missing this one. The implication for those who do miss it was left unspoken.

The mid-cap marketing equivalent of missing the AI revolution is not simply 'falling behind.' It is something more specific and more damaging: becoming colonized by the tools themselves.

The same marketing capability gaps that leak growth today are the entry points through which external forces — AI vendors, agencies, consultancies, platforms — could colonize the marketing function and reduce CMOs to mere tacticians. The CMO whose team lacks the strategic fundamentals to direct and govern AI does not stay at the same level of capability. She regresses. The tools do the pattern recognition, the media mix, the audience segmentation, the creative optimization — and the team becomes an operator of dashboards rather than a maker of decisions.

The mid-cap company that adopts AI without strategic fundamentals does not become more sovereign. It becomes more dependent — on whoever sold it the tool and controls the logic inside it. The AI vendor becomes the new holding company, the new agency group, the new external architect. Except this time the dependency is deeper, because the team does not even understand the logic of the decisions being made on its behalf.

III: Singapore can deploy at pace because it built the substrate

This is the part of the analogy that separates Singapore from countries of comparable size that are not generating 5% growth. Singapore spent decades building the non-negotiables: a world-class education system, English fluency, rule of law, clean governance, digital infrastructure, a civil service that could interpret and act on data. When AI arrived, it had a surface to land on.

Vivian himself made this explicit at the AI Engineer conference in Singapore in May: the country should focus on deploying AI widely, but deployment presupposes that the people doing the deploying understand what they are deploying and why. He warned against throwing every problem at a large language model and argued there is still a role for rule-based expert systems that simulate human knowledge. The human brain, he said, is often far more efficient than the energy-gobbling systems we have today.

Translate that to the marketing function and the operational principle is clear: you cannot deploy AI into a capability the team does not yet understand. You cannot in-house what the team cannot yet run. The sequencing is not AI first, capability second. It is capability first — or at the very least, capability alongside — so the team can govern the agents rather than be governed by them.

For a Fortune 500 company, this gap is survivable. They have the budget to course-correct, the headcount to absorb mistakes, the data volumes to eventually train models properly. For a Russell 2000 company with a marketing team of fifteen and a non-working budget already under pressure, the gap is existential and would often be fatal. They cannot afford to buy the wrong AI stack, discover eighteen months later that it has not moved the needle, and start over. Their version of 'missing the industrial revolution' is a lost three years of compounding growth leakage that a competitor with better fundamentals captured instead.

IV: The diagnostic question

The question Vivian was answering for Singapore — are we ready to deploy AI from a position of strength, or will AI deploy us? — is the same question every mid-cap CMO must answer for their marketing function. And the answer cannot be instinct. It has to be measured.

Does your team have the strategic fluency in the channels that drive your growth? Can they read first-party data and set audience strategy, or does someone else do that for them? Do they understand the logic of the algorithms they depend on, or are they optimising against a metric because a platform told them to? Can they brief, direct, and govern AI agents — or will the agents brief them?

These are not hypothetical questions. They are measurable conditions, and they determine whether AI will be a force multiplier or a force replacement.

The mid-cap CMO's version of that bet is the same: own the capabilities that drive your growth, deploy the tools that multiply their effect, and know — with evidence, not instinct — which is which.

The companies that do this will be the Singapore’s of the marketing landscape in the AI era : small, fast, fundamentally strong, and disproportionately successful.

The brands that do not will discover what Vivian already knows: if you are not at the table, you are on the menu.

Next
Next

The Personalization Industrial Complex