For many small businesses, idle cash is a quiet drag on the balance sheet. Money set aside for payroll, rent, suppliers or tax often sits in a corporate bank account earning little, because owners cannot afford to lock it away for months in search of a better return.

Chocolate Finance is now trying to turn that gap into its next growth line.

Also Read: Our main competition in Singapore is the idle cash lying in banks: Kristal.AI CEO

The Singapore-based cash management platform will officially launch Chocolate Business Accounts this week at The Business Show Asia 2026, marking its expansion from consumer savings-style products into corporate cash management. The new product is aimed at businesses that want to earn returns on spare operating cash while keeping access to funds when they need them.

Chocolate Business is not a corporate bank account and not a fixed deposit. It is structured as a cash managed account, meaning customer funds are placed into investment products with the aim of generating returns. That distinction matters: unlike a bank deposit, returns are not guaranteed, and capital can be exposed to investment risk. But unlike a fixed deposit, businesses are not tied to a set tenure.

The launch comes two years after Chocolate Finance introduced its consumer platform. Since then, the company says it has grown to about US$1.23 billion in assets under management and more than 150,000 customers.

The SME cash problem

The pitch is simple: businesses need liquidity, but liquidity often earns very little.

Chocolate Finance cited an industry survey estimating that SMEs in Singapore lose around US$616 million in potential interest each year by leaving idle cash in low-yield accounts. The same survey found that 45 per cent of SMEs identified liquidity as a priority.

That tension is familiar across Southeast Asia. Smaller companies often run on uneven cash cycles, waiting for client payments while still having to meet monthly payroll and supplier bills. In markets such as Singapore, where wage, rental and financing costs remain high, the ability to earn even modest incremental returns on unused working capital can make a difference.

At the same time, many SMEs lack the treasury teams that larger companies use to move surplus cash between money-market funds, short-term instruments and deposits. For a founder or finance manager handling day-to-day operations, the default option is often the easiest one: leave the cash in the bank.

Walter de Oude, Founder and CEO of Chocolate Finance, framed the product around that pain point. “Businesses need their cash available for payroll, suppliers and whatever comes next, but there’s no reason it should sit around earning next to nothing in the meantime,” he said.

How Chocolate Business works

Chocolate Business currently offers 1.5 per cent per annum on the first SGD300,000 (US$231,000), supported by the company’s Chocolate Top-Up Programme during a qualifying period. Amounts above that can earn up to 1.5 per cent per annum.

Businesses can fund their accounts through FAST transfers or PayNow. There is no minimum or maximum deposit requirement, according to the company. Withdrawals can be requested at any time, with no fixed tenure, withdrawal charges or penalties, although funds typically take one to two business days to arrive depending on the amount.

Business owners can monitor their balances and returns through a dedicated Chocolate Business app, available on the Apple App Store and Google Play Store.

Also Read: SEA’s digital payments boom has a dirty secret: SMEs still run on cash

Chocolate Finance also says it charges no upfront fees. Instead, it makes money only after meeting its target return. This model is designed to align the company’s incentives with customers, though the final outcome for users still depends on market conditions and the performance of the underlying cash management strategy.

The company is regulated by the Monetary Authority of Singapore and operates under Chocfin Pte Ltd. It is backed by Peak XV Partners, Prosus, Saison Capital and GFC.

Why now?

The move into business accounts is a natural extension for Chocolate Finance, but it also comes at a time when cash management is becoming more visible in Southeast Asia’s startup and SME ecosystem.

During the low-rate years, many companies gave little thought to short-term cash returns. That changed as global interest rates rose and founders began looking more closely at runway, treasury discipline and capital efficiency. Even as rate cycles shift, the habit of asking whether idle cash is being used properly is likely to stay.

Singapore is a particularly suitable testbed. The city-state has a high concentration of SMEs, a sophisticated financial infrastructure, broad use of PayNow and FAST, and a regulatory environment that has allowed digital wealth and cash management platforms to grow. For startups and SMEs operating across the region, Singapore often serves as both headquarters and treasury hub.

Still, Chocolate will need to tread carefully. Business owners are generally more conservative with operational cash than consumers are with spare savings. The money in a corporate account may be needed for salaries or urgent supplier payments, which leaves little room for confusion over risk, redemption timing or the difference between a managed account and a bank deposit.

That education challenge is likely to shape how quickly products such as Chocolate Business gain traction.

A crowded financial services layer

Chocolate Finance is not entering an empty market. Last year, Aspire, an all-in-one finance platform in Singapore, launched a new investment product designed to give SMEs institutional-grade returns on idle business funds without sacrificing liquidity or accessibility.

The competition will also come from incumbent banks, which still own the primary corporate account relationship and can bundle deposits with payments, lending and trade services.

It also sits adjacent to digital wealth and cash management players such as StashAway, Syfe and Endowus, which have helped popularise cash management products among retail users. Globally, business finance platforms such as Brex, Mercury, Wise Business and Revolut Business have shown how startups can build operating accounts and treasury tools around SMEs and technology companies, though their models and regulatory structures vary by market.

Also Read: Singaporean SMEs bleeding millions due to poor cash management

Chocolate’s challenge is therefore not only to offer a better return, but to become trusted enough to hold business operating cash. That is a higher bar than attracting consumer deposits for spare funds.

Beyond consumer finance

For Chocolate Finance, the launch signals a shift from a consumer-facing proposition into a broader cash management platform. If the product works, it could open up a larger pool of assets: SMEs typically hold more cash per account than individuals, and their balances can be more stable when tied to business operations.

But the business segment is also less forgiving. A delayed withdrawal or misunderstood product feature can have consequences beyond user inconvenience. For SMEs, cash flow is survival.

The larger question is whether Singapore’s SMEs are ready to treat idle cash as something that should be actively managed rather than simply parked. Chocolate Finance is betting that the answer is yes, and that businesses, like consumers, are beginning to expect more from the money sitting still.

The post Chocolate Finance targets idle SME cash with new business account appeared first on e27.

For a few years, Indonesian agritech was sold a seductive story: that enough venture capital could compress years of supply-chain building into a few funding cycles. Startups raised money to digitise farmers, connect harvests to buyers, extend credit, sell inputs and organise fragmented rural markets at speed. The pitch was familiar across Southeast Asia’s boom years: build the network first, figure out profitability later.

That story has now run into the physical reality of Indonesian agriculture.

According to the “AgTech Investment in Emerging Markets 2025” report prepared by AgBase, Briter, and Mercy Corps, the country’s smallholder farming system is not a clean consumer internet market waiting for an app. It is an archipelago of dispersed producers, uneven logistics, variable quality, limited cold-chain infrastructure, informal credit, and deeply local trading relationships. Treating farmers like conventional digital users — to be acquired, subsidised and retained through software alone — missed the core problem. Indonesian agriculture is not short of coordination tools as much as it is short of reliable operating rails.

Also Read: Agritech does not empower women farmers, until the system is fixed

The result is a sector-wide rethink. Since the capital tightening that began in 2022 and deepened through 2023, the conversation has shifted from growth-at-all-costs to operational discipline. In practical terms, that means fewer vanity metrics, more attention to unit economics, and a greater willingness to build or control the messy offline pieces that make agritech work.

From venture bets to system bets

The earlier agritech boom was shaped by abundant global capital and a high tolerance for risk. In that environment, many models were rewarded for expanding rapidly: signing up farmers, increasing gross merchandise value, and showing market share, even when every transaction needed subsidy support.

That approach was never unique to Indonesia. Across Southeast Asia, startups in logistics, fintech, commerce and food delivery went through similar cycles. But agriculture exposed the weakness more sharply because the underlying infrastructure gaps were harder to ignore. A subsidised digital marketplace may generate activity, but if the harvest cannot be graded, stored, financed, transported and sold reliably, the marketplace remains fragile.

The new phase looks different. Investors are increasingly backing “system bets” rather than speculative user-growth plays. These are companies that align with food security, supply-chain resilience, export readiness and corporate procurement needs. Instead of assuming software can replace complexity, they work within it.

Capital is also changing shape. Southeast Asia remains more equity-driven than grant-heavy ecosystems in parts of Africa, but investors are becoming more selective about where risk sits. Blended capital structures are gaining relevance: concessional funding can help absorb early infrastructure risk or finance expensive physical assets, while commercial equity can scale the parts of the business that have already been proven.

This is a meaningful shift. It recognises that some of the most important work in Indonesian agritech may not look like classic venture-backed software. It may involve field teams, warehouses, fulfilment centres, logistics partnerships, cold storage, quality-control systems and long-term relationships with processors or retailers.

The phygital reality

The most durable Indonesian agritech models are unlikely to be purely digital. They will be “phygital”: combining software with people and physical infrastructure.

Field agent networks are a good example. During the boom, offline teams were sometimes viewed as a drag on scalability. In practice, they are often essential. Agents help verify farm conditions, support credit underwriting, monitor crop quality, train farmers, and create the trust needed for repeat usage. In smallholder agriculture, a human layer is not merely customer support; it is a risk-management tool.

The same is true for hard infrastructure. Platforms that want to serve higher-value buyers need reliable standards. That requires control over grading, storage, aggregation and transport. Without this, a startup may facilitate transactions, but it cannot guarantee quality or traceability. For institutional buyers, especially processors, exporters and modern retailers, those guarantees matter.

This explains why “asset-light” models have lost some of their appeal. Being asset-light works when the market already has dependable infrastructure that a platform can plug into. In much of Indonesian agriculture, the infrastructure is incomplete. Startups either have to build parts of it, partner closely with those who own it, or accept limited control over their own service quality.

Also Read: Need of the hour: How agritech platforms can protect farmers from climate change

That makes the business harder, but also more defensible. A startup that can combine farmer relationships, operating infrastructure and data becomes more than a marketplace. It becomes a supply-chain partner.

The money is downstream

One of the clearest lessons from the reset is that smallholders cannot be the main payer for most digital services. Many farmers operate on thin margins and face volatile income. Charging them directly for apps, advisory tools or data products is often commercially unrealistic.

The stronger models are monetising downstream. They make money from processors, retailers, exporters, lenders, insurers or agribusinesses that benefit from better-quality supply, improved traceability, lower default risk or more predictable procurement. In this version of agritech, the farmer remains central, but the revenue pool sits closer to the buyer.

This is where impact and commercial returns can overlap. If a platform helps increase smallholder income by 20 to 30 per cent, it is not only producing a social benefit. It is also improving farmer loyalty and reducing repeated acquisition costs. If bundled agri-finance models — combining credit, insurance and guaranteed offtake — can keep repayment rates above 95 per cent, they show that farmer stability is directly linked to financial performance.

The point is not that impact automatically creates profit. It is that in agriculture, reducing risk for farmers often reduces risk for the platform too. Better income stability means better repayment. Better production standards mean better buyer retention. Better traceability means stronger access to corporate and export demand.

For Southeast Asia, this matters beyond Indonesia. Vietnam, the Philippines, Thailand and parts of Malaysia all face variations of the same challenge: fragmented production systems trying to serve increasingly formal, data-hungry and climate-conscious supply chains. The winners will not simply be the startups with the most downloads. They will be the ones that can translate farm-level activity into dependable commercial supply.

Building for exits, not headlines

The shift also changes how founders should think about exits. In Southeast Asia, the public-market path remains narrow for many startups, especially in specialised sectors such as agritech. Strategic mergers and acquisitions are a more realistic outcome.

That means companies need to be built with acquirers in mind. Regional agribusiness groups, food conglomerates, processors and input companies are unlikely to buy speculative growth alone. They will look for repeatable revenue, operational compatibility and technologies that improve their existing assets.

Farm management software, biological inputs, credit tools, traceability systems and procurement platforms all have potential value, but only if they fit into the workflows of large buyers. A startup that can be plugged into an agribusiness profit-and-loss statement has a clearer path to acquisition than one that only shows user growth without cash-flow discipline.

This is a colder, but healthier, market. It rewards founders who understand procurement cycles, logistics costs, repayment behaviour and quality assurance. It is less forgiving of companies that hide weak economics behind GMV.

Also Read: The agritech challenge in Indonesia: Can AI and mobile apps enhance productivity?

Indonesia’s agritech correction should not be read as a verdict against the sector. The country still has enormous agricultural complexity to solve, and that complexity creates room for valuable companies. But the old playbook has expired.

The next generation of Indonesian agritech will be less glamorous and more operational. It will blend software with fieldwork, capital with infrastructure, and farmer impact with downstream commercial demand. In other words, it will stop trying to leapfrog the hard parts of agriculture — and start building through them.

The post Why Indonesia’s agritech winners will be phygital, not purely digital appeared first on e27.

Southeast Asia’s electric vehicle (EV) story is no longer just about flashy car launches or government targets. Increasingly, the region’s EV transition is being shaped by a younger group of startups tackling less glamorous but commercially urgent problems: motorbike electrification, fleet charging, battery swapping, electric ferries, logistics vehicles and mobility services.

A new ranking by private market intelligence platform Tracxn points to how quickly this segment has moved from climate-tech promise to venture-backed experimentation. According to its August 2026 report, the top 16 funded EV startups across Singapore, Indonesia, Thailand and Vietnam have raised a combined US$622 million in equity funding. Every company in the cohort was founded in 2016 or later.

Also Read: Can Southeast Asia power the EV and chip boom without leaving communities behind?

The list covers companies with at least US$5 million in disclosed cumulative equity funding. That threshold leaves out many early experiments, but it gives a useful view of which models have managed to convince investors that EV adoption in Southeast Asia is not a distant policy ambition, but a market being built now.

Singapore leads the funding table, but not necessarily the roads

Singapore accounts for half of the cohort, with eight of the 16 companies based in the city-state. That may seem counterintuitive. Singapore has a small domestic vehicle market, limited land for large-scale manufacturing and strict rules around vehicle ownership. But for EV startups, the country plays a different role: it is a financing, headquarters and regional expansion base.

Startups based in Singapore can build corporate teams, raise from international funds, access regional customers and structure operations across multiple Southeast Asian markets. For EV companies, that is particularly important because the business is rarely confined to one activity. A startup may design hardware in one country, manufacture through partners in another, sell into logistics fleets across the region and rely on software to manage charging, batteries or financing.

Indonesia, with four companies in the cohort, represents the other side of the equation. It is Southeast Asia’s largest automotive market and home to one of the world’s most important nickel reserves, a key material for many EV batteries. The country has been trying to move up the battery and EV value chain, attracting global manufacturers while also encouraging local adoption of electric two-wheelers and buses.

Thailand and Vietnam each have two companies in Tracxn’s list. Thailand has long been the region’s automotive manufacturing hub and has set out ambitions to make EVs a significant share of production in the coming years. Vietnam, meanwhile, has already produced one of Southeast Asia’s most visible EV names in VinFast, though Tracxn’s ranking focuses on startups rather than the broader industrial champions reshaping the market.

Why two-wheelers matter more than cars

For Southeast Asia, the EV opportunity cannot be understood through a US or European lens, where electric cars dominate the discussion. In much of the region, the motorbike is the everyday vehicle. It is used for commuting, food delivery, courier work, informal trade and last-mile logistics. That makes electric two-wheelers one of the most practical routes to cutting fuel costs and urban pollution.

The economics are compelling, but not simple. Electric motorbikes can have lower running costs than petrol models, yet upfront prices, battery reliability, resale value and access to charging remain major barriers. This is where startups are trying to find a wedge. Some focus on battery-as-a-service, allowing users to rent or swap batteries rather than own them. Others target delivery fleets, where predictable routes and high daily mileage can make electrification easier to justify.

Fleet customers are especially important. A consumer may hesitate over an electric motorbike if charging is inconvenient or resale prices are uncertain. A logistics company, ride-hailing partner or food delivery operator can make a more data-driven decision, calculating fuel savings, maintenance costs and vehicle downtime across hundreds or thousands of units.

Also Read: 70 Cars, 55,000 hours, one question: Can you forecast when an EV battery will die?

That explains why Tracxn’s cohort spans not only vehicle makers, but also companies in charging infrastructure, mobility-as-a-service and commercial EVs. The region’s EV transition is less about one product replacing another and more about an ecosystem forming around energy, hardware, software and financing.

Funding is maturing, but still selective

The companies in Tracxn’s ranking range from seed stage to Series B, with most sitting around Series A or Series B. This suggests a market that has moved beyond early pilots but has not yet reached the maturity of fintech, e-commerce or logistics software in Southeast Asia.

That matters because EV startups are capital-intensive. Unlike pure software companies, they often need to deal with hardware design, inventory, servicing networks, regulatory approvals and physical infrastructure. Even charging software companies eventually run into real-world constraints: grid capacity, property access, utilisation rates and the economics of installing equipment before demand is fully proven.

The investor mix reflects that complexity. Tracxn notes that backers of companies in the cohort include Peak XV Partners, Jungle Ventures, GSR Ventures, TVS Motor and Horizons Ventures. The presence of both venture capital firms and corporate investors is telling. Financial investors are looking for scalable models in a large emerging market. Strategic investors, including automotive and mobility-linked players, are watching for technologies, distribution models or local operators that could shape future demand.

Still, US$622 million across 16 companies is modest when compared with the billions poured into EV and battery companies in China, the US and Europe over the past decade. That may not be a weakness. Southeast Asia’s EV market is fragmented by regulation, income levels, grid readiness and consumer behaviour. The winners are unlikely to be those that simply copy global EV playbooks. They will be companies that adapt to dense cities, cash-sensitive consumers, informal transport networks and fleet-heavy usage.

Policy is pulling the market forward

Government policy remains a major force. Singapore has said it wants to phase out internal combustion engine vehicles by 2040. Thailand has offered incentives to attract EV production and stimulate local demand. Indonesia has used its nickel resources as leverage to build a battery and EV manufacturing base. Vietnam has paired domestic industrial ambition with growing consumer awareness of electric mobility.

But policy alone cannot build adoption. Subsidies can lower prices, but they do not solve charging anxiety. Manufacturing incentives can bring factories, but they do not guarantee affordable financing or reliable after-sales service. This is where startups can matter: they often operate in the messy gaps between public ambition and consumer behaviour.

The next test is scale. Many EV startups can run pilots, sign memoranda of understanding or deploy small fleets. Fewer can prove that customers will pay consistently, batteries will last as promised, utilisation rates will support infrastructure costs and maintenance can be handled across cities and islands.

Also Read: Exponent Energy unlocks a zero to 100 per cent 15-min rapid charge for electric vehicles

Tracxn’s ranking is therefore less a victory lap than a progress marker. Southeast Asia now has a visible group of EV startups with meaningful investor backing. The harder question is which of them can turn that funding into vehicles on roads, batteries in circulation, chargers that are actually used and business models that survive without permanent subsidy.

For a region where transport demand is still rising and urban air quality remains a daily concern, that question is not academic. Clean mobility in Southeast Asia will not be delivered by carmakers alone. It will be built through a patchwork of two-wheelers, fleets, ferries, batteries, chargers and software — and the startups now attracting capital are beginning to show what that patchwork might look like.

The post Southeast Asia’s EV startups draw US$622M as clean mobility shifts from pitch to pilot appeared first on e27.

The cost of AI is not only hidden in data centres. It is hidden in the systems we are learning to depend on.

Artificial intelligence still feels cheap. A manager pays for a monthly subscription. A developer opens a coding assistant. A student asks a chatbot to summarise a report. The interaction is instant, polished and priced in a way that makes the cost feel almost trivial.

That impression is unlikely to last.

The first phase of consumer AI was built around the prompt: one user, one question, one answer. The next phase is being built around agents. These systems do more than respond. They search, plan, call tools, inspect files, write code, check their work, retry failed steps and continue until a task is complete.

That shift changes the economics. The relevant unit is no longer the prompt. It is the workflow.

This matters because much of the public debate still treats AI as a software product. In practice, it is becoming an infrastructure layer. It depends on data centres, electricity, water, chips, cloud providers, model vendors, monitoring systems and compliance processes. As adoption spreads, the cost moves beyond the subscription price.

The early signs are already visible.

GitHub has said Copilot will move to usage-based billing from June 2026. Instead of counting premium requests, usage will be calculated through AI credits linked to token consumption, including input, output and cached tokens. Paid plans will include a monthly allowance, with additional usage billed separately.

For users, this may look like a pricing update. For the market, it is a signal. AI is beginning to move away from the flat-rate psychology of software-as-a-service and toward the metered economics of cloud infrastructure.

A short chatbot reply and a long coding-agent session are not the same thing. The latter may involve multiple model calls, large context windows, file analysis, code edits and repeated attempts to solve the problem. The user sees one task. The system may perform hundreds of computational steps behind the interface.

That is where the idea of cheap AI starts to fray.

The subsidy phase

The current AI market has many features of a subsidy phase. Large technology companies are competing for market share, user habits and strategic position. Prices are simple because simple pricing accelerates adoption. Free tiers reduce friction. Flat-rate plans make experimentation feel low-risk. Enterprises are encouraged to deploy before they fully understand the cost of scaled usage.

There is nothing unusual about this. Search, cloud storage, ride-hailing and food delivery all used cheap access to change behaviour before their economics became clearer. AI may follow the same pattern, but with a much heavier infrastructure burden.

The counterargument deserves serious attention. AI is becoming cheaper at the unit level. Model inference costs have fallen sharply. Hardware is improving. Smaller models are becoming more capable. Energy efficiency is rising. A narrow claim that AI costs can only go up would be wrong.

But lower unit costs do not guarantee lower total costs. When technology becomes cheaper and easier to use, demand often expands. Once AI is embedded into office software, coding tools, customer support, finance, logistics, marketing, compliance and public administration, total consumption can grow faster than efficiency gains.

The important issue is not whether one query becomes cheaper. It probably will. The issue is whether aggregate AI demand grows faster than the systems built to support it.

Also Read: The transformation ecology crisis: How AI is exposing the hidden fragility of high-performing teams

The data centre bill

The International Energy Agency expects global electricity consumption from data centres to more than double by 2030, reaching about 945 terawatt-hours. That would represent just under 3 per cent of global electricity consumption. From 2024 to 2030, data centre power demand is expected to grow at roughly 15 per cent a year, more than four times the rate of demand growth in other sectors.

AI is one of the main drivers. In the United States, data centres are expected to account for nearly half of electricity demand growth between now and 2030.

Those figures do not mean AI will overwhelm power grids everywhere. They do show that AI is no longer only a software story. It is becoming a major industrial load.

That distinction matters. Software scales quickly. Power systems do not. Data centres can be planned, financed and built faster than new transmission lines, generation assets and regulatory approvals can be delivered. The North American Electric Reliability Corporation has warned that large data centre loads can arrive faster than the infrastructure needed to serve them.

The mismatch is obvious. AI companies move at software speed. Grids move at infrastructure speed.

This is where the cost may move outward. If utilities need to build new generation, strengthen transmission and reserve firm capacity for data centres, someone has to pay. In some cases, the technology companies will bear much of that burden. In others, the cost may be spread through electricity rates, tax incentives, public investment or delayed infrastructure priorities.

The electricity bill for AI will not always appear on the AI invoice.

From convenience to dependency

The larger cost may be structural.

AI is already being considered or deployed in banking, identity verification, cybersecurity, fraud detection, public administration, healthcare operations, logistics and infrastructure management. In those settings, the risk is not limited to a chatbot producing a poor answer. The risk is that institutions redesign important processes around systems they do not fully control.

The Financial Stability Board has warned that AI use in finance can create vulnerabilities linked to third-party dependency, service-provider concentration, cyber risk, model risk, data quality and governance. These are not abstract concerns. Modern AI depends on a narrow stack of model providers, cloud platforms, chip suppliers, data pipelines and application layers.

Once AI is embedded into a critical workflow, the institution becomes dependent on that stack. Each layer can fail. Each layer can change its pricing. Each layer may sit outside the direct control of the organisation relying on it.

This dependency is manageable when AI supports an existing human process. It becomes more serious when AI replaces the process.

A bank that uses AI to help analysts review suspicious transactions still has a human workflow. A bank that restructures its compliance function around automated agents has a different risk profile. If the model changes, the vendor fails, the API goes down, regulation shifts, compute prices rise or power supply tightens, the bank may discover that the previous manual capability no longer exists.

The failure scenario is not only that AI breaks. It is that the fallback has disappeared.

Also Read: Why Malaysia’s AI Nation 2030 plan matters for B2B startups

The loss of institutional muscle

The labour debate around AI is often framed around jobs lost or productivity gained. That framing is too narrow. A more subtle risk is the erosion of competence.

Organisations learn through repetition. Junior staff become useful by drafting, checking, reconciling, researching, coding, correcting mistakes and handling exceptions. They build judgement because they spend time with the work.

If agents absorb too many of those tasks, companies may appear more efficient in the short term while weakening their own talent pipeline. The organisation keeps senior decision-makers and automated systems, but loses the layer of people who understand how the work is actually done.

This is difficult to measure. It would be careless to claim that AI has already hollowed out institutional competence across the economy. But the risk is credible, especially in fields where errors are detected through experience: finance, law, medicine, engineering, cybersecurity, aviation and public administration.

The danger is not that AI makes people incapable. The danger is that organisations stop creating people who know how to operate without it.

The cost of no return

The most expensive technology choices are not always the ones with the largest upfront price. They are the ones that remove optionality.

A company can test AI safely when it remains a layer on top of an existing process. The risk changes when the process itself is rebuilt. Teams are reduced. Vendor contracts are signed. Data flows are reorganised. Interfaces are redesigned. Compliance procedures are rewritten. Management dashboards begin to assume AI availability. New employees are trained on the AI-native workflow rather than the older one.

At that point, stepping back is no longer a matter of cancelling a subscription. The company would need to rebuild skills, processes, documentation, software and confidence.

This is one of the least visible costs of AI adoption. The bill is not only for compute. It is for irreversibility.

The counter case

A serious assessment must recognise that AI can also reduce costs. It can improve demand forecasting, detect fraud, support grid management, accelerate software development, strengthen customer service, optimise logistics and support scientific research. In some areas, AI may help reduce emissions or make infrastructure more efficient.

Nor are all AI systems equal. A small model running locally for a narrow task is not comparable to a frontier model powering an autonomous agent across multiple systems. A document summariser is not comparable to an always-on compliance, trading or cybersecurity agent.

The distinction matters. AI as a tool is one thing. AI as a dependency is another.

The answer is not to reject AI. That would be unrealistic and, in many cases, commercially damaging. The answer is to treat AI as infrastructure when it functions as infrastructure.

Also Read: The builders of AI are selling a future their own technology destroys

What should change

Companies need to measure AI by workflow cost, not subscription cost. A monthly plan tells management little about the economics of large-scale agent deployment. The relevant calculation includes compute, cloud services, monitoring, human review, compliance, security, error correction and fallback capacity.

Critical institutions also need continuity plans. Banks, public agencies, healthcare operators and infrastructure providers should be able to explain what happens if a model provider, API, cloud region, vendor contract or power supply becomes unavailable.

Regulators need better visibility into concentration risk. If many institutions rely on the same small group of model providers and cloud platforms, the exposure becomes systemic. The Financial Stability Board’s warnings on third-party dependency should be treated as an early signal, not a theoretical footnote.

Companies should also protect human competence. Keeping people “in the loop” cannot mean asking them to rubber-stamp machine output. It means preserving the ability to challenge, audit and replace AI-driven work. It also means continuing to train junior staff on fundamentals, even when automation appears faster.

Finally, AI infrastructure should face the same scrutiny as other strategic infrastructure. Data centres require power, water, land, transmission capacity and political permission. If the gains are private while the infrastructure costs are spread more widely, the public bargain will eventually be questioned.

The real price

The first phase of AI was about access. The next phase will be about dependency.

Cheap AI has encouraged experimentation, and much of that experimentation has value. But cheap access can hide expensive commitments. Once agents are embedded into the machinery of business and government, the cost of AI will no longer be measured only in tokens, subscriptions or data centre bills.

It will be measured in grid pressure, vendor concentration, institutional fragility, lost human capability and reduced room for reversal.

AI can produce useful work. That point is settled. The harder question is whether companies and governments are building around it with a clear view of what happens when it becomes more expensive, less available, less reliable or too deeply embedded to remove.

—

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.

Image credit: Tima Miroshnichenko

The post The true cost of AI is beginning to surface appeared first on e27.

On 8 April 2026, a ceasefire ended a regional conflict that had, for the first time, reached all six Gulf Cooperation Council states. The UAE was among the most affected. For a stretch in early March, Emirates and Etihad were running repatriation flights, and Gulf airspace was operating under precautionary measures.

Six weeks after that ceasefire, I moved my operating base there.

Here is what the market looked like when I did it. Between December 2025 and May 2026, Dubai property transaction values fell 55 per cent. In the DIFC — the financial district, the part institutional money buys first and sells last — they fell 67 per cent.

I signed in May.

People I respect asked whether I had lost my mind. I had left the safest hub in Asia — rule of law, a world-class regulator, a top-five global financial centre — for a region that had spent the spring on every front page.

They were asking the wrong question.

The question nobody asked me

If the safe market is so obviously good, why is everyone already in it?

Every investor reading this knows the rule. You buy in fear. You sell in greed. You get paid for holding what other people cannot stomach holding. We repeat it about equities. We repeat it about crypto. Then we build our expansion maps as though the rule stops at the border.

It does not. And it works better on geography than on assets, because unlike a stock, a country cannot be bid back up in an afternoon. The mispricing lasts for years.

One thing worth saying plainly before I go further, because I am Singaporean and this is my home market: crowded is not an insult. A market gets crowded because it is good. Everything I am about to describe is a statement about entry price and competition, not about quality — and the crowd is usually right about where the quality is. It is just early, and you are usually late.

What the fear actually did to the price

Not a feeling. Receipts.

Capital contracted across the region. MENA startups raised US$1.7 billion across 242 rounds in the first half of 2026, down 18 per cent year-on-year, with deal volume down 28 per cent. Analysts attributed a 22 per cent drag directly to the conflict. That is the fear, measured.

Now look at where the money that stayed actually went.

The UAE took US$1.2 billion of that US$1.7 billion — roughly 71 per cent of everything raised across the entire region — across 83 deals. Saudi Arabia, the region’s other giant, managed US$259 million across 80 deals, and not a single later-stage round. The UAE’s largest funded sector was fintech: US$409 million across 20 deals.

Read that as an operator rather than a reader. Capital fled the region, the survivors concentrated into one city, and the sector they concentrated into was mine. The competition thinned and the buyers stayed. That is the configuration you spend a career waiting for, and it lasted about a quarter.

Also Read: Your founder brand could add or subtract US$500K to US$1M before you walk into a room: Here’s how

Meanwhile the structural base never moved. The UAE closed 2025 with record foreign direct investment of AED 177.3 billion — a fourth consecutive record year, ninth in the world, on a total stock of AED 1.171 trillion. The Middle East led global greenfield capital expenditure growth at 72.4 per cent, with the UAE alone contributing 38 per cent of the region’s greenfield capex.

And the discount showed up in the price, not the fundamentals. While transaction volumes collapsed 55 per cent, DIFC prices still rose 19.3 per cent year-on-year. Dubai rents softened 6.2 per cent in the second quarter while home prices held above 2025 levels.

Volume fled. Value did not. Panicked sellers, intact asset — that is the textbook definition of buying in fear, and it was sitting in public data the whole time.

What the stress test proved

Three things became obvious this year that would never have become obvious in a calm one.

• The bubble was fake. A great deal of what passed for a business in 2024 and 2025 was narrative, financed by cheap money and momentum. When conditions turned, those companies did not slow down. They vanished. Nothing exposes a company faster than a quarter in which nobody is buying the story — and the stress test did in ninety days what most investors need two years of diligence to establish. If you were wondering which businesses in this region were real, you no longer have to wonder. The list is short and it is public.
• Stability is not the absence of threat. It is the ability to keep operating through one. This is the part I did not expect, and it is the reason I stopped hesitating. With the threat directly overhead, the city kept working. Flights resumed within days. Banks settled. Courts sat. Contracts were honoured. The currency peg held and nobody reached for capital controls. In the same window I watched businesses in far calmer parts of the world — the UK, several European markets — lose more operating days to their own policy cycles and domestic disruption than we lost to a regional conflict. One of those is a headline risk. The other is a structural one. Only the first shows up in a risk report, and it is the second that actually costs you a year.
• The noise left. This is the one founders should care about most. The tourist capital went home. The consultants who arrive for a boom went with it. What remained was the people who actually build — and that changes who you are competing with for attention. When the room empties, the institutions still deploying can finally see who is serious. Access I could not get in two years of a crowded market, I have had in the last four months, not despite the disruption but because of it. That 71 per cent concentration figure is this same fact viewed from the outside: the capital did not lose interest in the region. It lost interest in the noise.

Why this happens, every time

Seven things I would tell a founder before they write off a market because of a headline.

• Stability is priced in — you just don’t pay in cash. You pay in competition, in acquisition cost, in valuations set by twelve other funded companies solving your problem. A safe market is expensive the way a crowded trade is expensive: the price already reflects everything good about it.
• The risk premium is compensation, not punishment. Markets pay you to hold what others won’t. That is the foundation of asset pricing, and it applies to licences, partnerships, talent and equity exactly as it applies to bonds.
• Volatility is not risk. Risk is permanent loss — a rule change that kills your model, a licence you never get, a partner who takes your customers. Volatility is discomfort: headlines, a bad quarter, your mother calling to ask if you are safe. Most founders avoid discomfort and file it under risk management. The gap between the two is where the returns live.
• Competition thins exactly when the headlines are worst. The month a region leads the news is the month your competitors’ investment committees say “let’s revisit next year.” Fewer bidders, better terms. That 71 per cent concentration figure is what thinning looks like in a dataset.
• Instability removes everything fake. Weak balance sheets leave. Tourist competitors leave. A stressed market runs your competitive analysis for you, free, in about ninety days.
• Perception lags reality by years, and the lag is the arbitrage. A region gets described by its worst month for the following five. Everything in this article will be consensus by 2029 and worth almost nothing to act on.
• The broken plumbing is the product. A market where money moves in four days at eight per cent is not a warning sign. It is the business, sitting there, unbuilt.

Also Read: Why the smartest founders are interviewing investors before investors interview them

How to price a market yourself

You do not need my conviction. You need a method. Four checks, all runnable in an afternoon on public data.

• Split volume from price. Pull transaction counts and price indices separately for the same window. If volume is collapsing while prices hold or rise, you are looking at a liquidity event, not a value event — sellers are leaving, buyers are not. If both fall together, that is a genuine repricing and you should wait. Dubai in the first half of 2026 was emphatically the first case: volumes down 55 per cent, DIFC prices up 19.3 per cent. That divergence is the single most useful number in this article.
• Check concentration, not totals. A regional funding headline tells you almost nothing. Break it by city and by stage. A region where one hub takes 71 per cent of the capital, and where the second-largest market records no later-stage rounds at all, has one real destination regardless of what the map suggests. Totals describe a region. Concentration describes where you should actually be standing.
• Follow the corridor, not the country. Model where money physically moves — remittance flows, trade lanes, settlement routes — and ignore GDP rankings entirely. A country’s economy tells you how big it is. Its corridors tell you where the fees are, and fees are the only thing you can build a company on.
• Count what is disappearing, not only what is growing. Correspondent banking down roughly 30 per cent while transaction volume climbs is not a statistic. It is an infrastructure vacuum, and every vacuum is somebody’s business. Growth attracts competitors. Withdrawal creates openings. Most founders only ever screen for the first.

If three of those four point the same way, the headline is not describing the opportunity. It is describing the entry price.

Buy in fear is not buy blind

This is the part left out of every “go where it’s hard” article, and it is why most people who quote the principle lose money with it.

Buying fear only works with position sizing. The investors who blow up are not the ones who bought fear. They are the ones who bought it with everything they had and no way out.

Five rules I operate by:

• Keep the boring things boring. Legal domicile, treaty coverage, dispute resolution and custody stay somewhere the rule of law is not a variable. Take risk on the market, never on the courts.
• Separate your four bases. Domicile, licensing, operations and capital are four decisions, not one address. Most founders collapse them into one because that is how a company formation agent sells it.
• Be able to leave in 48 hours. Not because you plan to, but because knowing you can is what lets you commit properly while you are there.
• Never bet what you cannot lose twice. One market failing should cost you a quarter, not the company.
• Name what would actually end you, before you go. Not the scary thing — the terminal thing. If you cannot write it down, you have not done the work, and you are not buying fear. You are gambling with a good story attached.

Do that, and the trade stops being brave and starts being arithmetic.

The Gulf is the entrance, not the destination

The reason this matters to a Southeast Asian founder has nothing to do with Dubai as a lifestyle decision.

More than US$80 billion a year already moves from the GCC into India, the Philippines and Pakistan. India alone took a record US$129.4 billion in remittances in 2024, roughly 38 per cent of it from the Gulf. Pakistan booked a record US$41.6 billion in its last financial year.

Meanwhile the infrastructure carrying that money is disappearing. Correspondent banking relationships fell roughly 30 per cent between 2011 and 2022. Volume up, plumbing down. That gap is the business.

And the demand sits precisely where the founders are not. Some 1.3 billion adults remain outside the formal financial system, and 650 million of them live in eight countries — exactly one of which is in Southeast Asia. Around a quarter of Pakistani adults hold a bank account, against 56 per cent in Indonesia and 89 per cent in India.

For contrast, and I say this as someone who built here and still builds here: Southeast Asian startups raised US$1.85 billion across 229 transactions in the first half of 2025, a six-year low, with seed funding halving to US$50.7 million. That is a cycle, not a verdict. But cycles are precisely the thing you are supposed to trade, and most of us don’t.

Also Read: Almost got “digitally arrested.” India needs Singapore’s playbook before the next scam call

I had already been taught this, somewhere else entirely

I gave a TEDx talk this year about leadership forged in darkness. The argument was that the capability you actually rely on is never built in calm conditions. It is built in the ones that strip everything non-essential away. Instability removes everything fake.

I was talking about people. It took me embarrassingly long to notice it is the same sentence about markets — and that this year simply proved it at scale.

A comfortable market lets you fake product-market fit. Cheap capital and working infrastructure will keep a mediocre product alive long enough for you to mistake a funding round for traction. A market under stress offers no such mercy.

Operating where the infrastructure is missing taught my teams things a clean market never could: how to build while the regulator is still writing the rules, how to run when settlement fails at 2am, how to earn trust from people whose institutions have failed them their whole lives. Those are not hardships to endure on the way somewhere better. They are the apprenticeship, and they are what makes a company hard to copy once the market matures.

Founders keep asking me which market is easiest to enter. Ease of entry is a warning, not a feature. It tells you exactly how low the barrier will be for whoever comes after you.

What Singapore does better than anywhere

I am Singaporean. I am not writing this to run my own country down, and nothing above should be read that way.

Rule of law. Contract enforcement. A regulator in MAS that builds alongside founders rather than merely supervising them. Treaty coverage. Depth of financial talent. And the plain fact that a Singapore entity opens doors that entities from almost anywhere else have to knock twice for. Those are not small advantages. They are the reason so much of the region’s serious capital is structured here, and they are why I still am.

I hold structure, licences and relationships there, deliberately. Singapore is where I keep the things that must never be volatile — and in a year like this one, that turned out to be worth more, not less.

The mistake was never choosing Singapore. It was assuming my headquarters and my centre of gravity had to be the same city. They are two different jobs, and only one of them has to be where the growth is.

The trade

The safest market on your slide is safe because it is finished. Everything good about it is known, priced, and already being competed for by people who arrived before you.

Fear is not a signal to stay away. It is a signal that the price is temporarily wrong, and that most of the people who should be bidding against you are currently doing something else.

Transactions fell 55 per cent. Prices rose 19. Somebody was selling.

Buy in fear. It works for portfolios. It works for maps.

—

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 Your safest market is a crowded trade: Why I moved to Dubai six weeks after the ceasefire appeared first on e27.