For years, artificial intelligence was framed as a technology for companies with deep pockets: banks with large data teams, manufacturers with automated lines, or global platforms sitting on oceans of customer information. Malaysia’s latest AI agenda is trying to challenge that assumption.

Under the National AI Action Plan 2026-2030, also known as AI Nation 2030, the government is positioning AI not only as a tool for frontier industries, but as basic economic infrastructure for the sectors that keep the country running: micro, small, and medium enterprises, farmers, and plantation smallholders.

Also Read: Malaysia’s sovereign AI bet: Local context becomes the next startup moat

That matters because these groups are often the least equipped to adopt advanced technology, even as they have the most to gain from it. MSMEs dominate Malaysia’s business landscape, accounting for 84.4 per cent of all businesses in the services sector. In agriculture and plantations, smallholders and traditional producers remain critical to food supply, rural employment, and export-linked value chains such as palm oil.

The plan’s central bet is simple: AI adoption will not spread widely if it depends on every small firm or farmer building their own systems from scratch. Instead, Malaysia wants to lower the barrier to entry through shared platforms, vetted tools, common datasets, and training programmes that make AI usable without requiring every user to become a technologist.

From digitalisation to AI adoption

For many Malaysian MSMEs, the problem is not a lack of interest in technology. It is a lack of time, skills, and clarity.

A small retailer, logistics operator, home services provider, or food business may already use digital payments, accounting software, or online marketplaces. But moving from basic digitalisation to AI-enabled operations is a larger step. It requires knowing which tools are reliable, how they connect to existing workflows, and whether the benefits justify the cost.

The AI for MSMEs Impact Engine, labelled I10 in the plan, is designed to address this gap. Building on the Business Digitalisation Initiative, it aims to give MSMEs structured access to AI through a one-stop enablement ecosystem. Rather than asking small business owners to navigate a fragmented market of software vendors, the plan calls for modular and pre-vetted AI tools that can be integrated into platforms they already use.

The practical applications are not hard to imagine. AI can help a small retailer forecast demand, automate inventory tracking, answer customer queries, generate marketing content, or streamline invoices and payments. For a services business, it can support appointment scheduling, document processing, customer segmentation, and internal reporting.

The ambition is to provide 1.5 million MSMEs with scalable access to AI. If executed well, that could shift AI from being a premium productivity layer for larger companies into a utility for everyday businesses.

This is also where Malaysia’s plan fits into a wider Southeast Asian challenge. Across the region, MSMEs employ large numbers of people but often struggle with thin margins, low productivity, and limited access to digital talent. Governments from Singapore to Indonesia have launched digital adoption programmes, but AI introduces a new policy question: how to make advanced tools affordable and trustworthy for businesses that cannot absorb costly failed experiments.

Also Read: Why digitalising SMEs matters for Southeast Asia’s economic resilience

Malaysia’s answer is to create an AI marketplace where local providers can offer vetted, affordable services. This could also support domestic AI startups by giving them a clearer route to serve smaller customers at scale.

Bringing precision farming to small producers

The same logic runs through the plan’s approach to agriculture. AI in farming is often associated with large commercial operations using drones, satellite data, automated irrigation, and predictive models. Malaysia wants to make those capabilities available to smaller producers too.

The Agrofood: Scalable Agristack initiative, or I6, focuses on using data to improve precision farming and predictive analytics. In practice, this means helping farmers make better decisions about when to irrigate, how much fertiliser to apply, and how to reduce losses from pests, disease, and climate volatility.

The early phase will begin with pilots for precision irrigation and fertilisation in selected paddy and vegetable clusters. The plan then expands into weather analytics, automated pest detection, and a wider range of crops, including fruits.

This is not just about efficiency. Food security has become a sharper concern across Southeast Asia as countries deal with volatile commodity prices, changing weather patterns, and pressure on arable land. Malaysia, like many of its neighbours, must balance import dependence with the need to strengthen domestic production.

A scalable agristack gives the government and producers a shared digital architecture for agricultural data. If built carefully, it can allow farmers who lack expensive private systems to benefit from common datasets and AI models. That could help move decision-making from instinct alone to a mix of local experience and predictive insight.

The challenge will be trust. Farmers will not adopt AI simply because a platform exists. Tools must work in local languages, reflect local crop conditions, and prove their value in the field. Extension officers, cooperatives, universities, and agritech startups will likely play a crucial role in turning national infrastructure into everyday adoption.

Palm oil smallholders and the data divide

Malaysia’s plantation sector faces a similar divide. In palm oil, smallholders manage about 26.4 per cent of the country’s planted area. Yet they often operate with far less capital, data access, and technical support than large estates.

The AI Platform for Smallholder Empowerment, or I9, aims to narrow this gap by consolidating datasets into a unified platform such as MySawit. The goal is to give smallholders access to tools for pest and disease management, fertiliser optimisation, and more efficient monitoring.

The plan projects up to a 70 per cent reduction in manual labour and up to 2.7 times greater land coverage through AI-enabled automation, including drone services for spraying and monitoring. If those gains materialise, they could help smallholders improve yields and the quality of fresh fruit bunches, while reducing dependence on labour-intensive fieldwork.

The palm oil industry is also under growing scrutiny from global buyers and regulators over sustainability, traceability, and land-use practices. Better data systems could therefore serve a dual purpose: raising productivity for smallholders while helping the sector respond to market demands for transparency.

The missing pieces: data, skills, and inclusion

AI systems are only as useful as the data and people behind them. AI Nation 2030 recognises this through enabling initiatives such as the AI-Ready Data Ecosystem, which seeks to aggregate priority datasets and make them available through a National Data Exchange.

The proposal for a “Right-to-Data” channel for non-sensitive public sector data is particularly important. Local AI developers and agritech startups need access to reliable datasets to build models suited to Malaysian conditions, rather than depending entirely on generic imported tools.

Talent is the other foundation. The Talent Pipeline @ Scale initiative includes individual-based AI skilling credits aimed at workers in at-risk professions and low-income groups. This is a necessary safeguard. If AI adoption is framed only as automation, it will create anxiety among workers. If it is linked to upskilling and productivity gains, it has a better chance of being seen as augmentation rather than replacement.

Malaysia’s human-centric framing, tied to the MADANI vision, gives the plan its political and social logic. The test, however, will be implementation. Inclusive AI requires more than national targets. It needs simple procurement channels, local support networks, affordable tools, and measurable outcomes for users who do not have the luxury of experimenting endlessly.

Also Read: Meet the Malaysian AI startups pushing beyond the ChatGPT hype

By putting MSMEs, farmers, and smallholders near the centre of its AI strategy, Malaysia is making a statement about where digital transformation should happen next. The country’s AI future will not be judged only by the sophistication of its research labs or the scale of its data centres. It will also be judged by whether a paddy farmer in Kedah, a palm oil smallholder in Sabah, or a neighbourhood retailer in Johor can use AI to make better decisions and earn more from their work.

If AI Nation 2030 delivers on that promise, Malaysia could offer Southeast Asia a useful model: one where artificial intelligence is not just a race for the most advanced firms, but a practical tool for lifting the economic floor.

The post From paddy fields to small shops, Malaysia maps an inclusive AI future appeared first on e27.

Chinese enterprises are increasingly setting their sights on Southeast Asia, and the reasons go well beyond the familiar “supply chain diversification” narrative. e27 speaks to Lisa Li, China Lead Partner, KPMG Global China Practice, to unpack what is genuinely driving capital allocation decisions in Chinese boardrooms today.

From market opportunity to boardroom decision

According to Li, Chinese business leaders are turning to Southeast Asia due to a combination of pull factors, push factors, and strategic alignment.

The region’s enormous consumption potential — fuelled by a young demographic, rising disposable incomes, and rapid urbanisation — makes it an attractive frontier for Chinese companies in consumer goods, e-commerce, and digital services looking to expand their customer base.

Cost also plays a role. Land, labour, and utilities remain comparatively cheaper across most Southeast Asian countries than in China’s coastal manufacturing hubs, allowing companies to protect margins while diversifying production.

Also Read: Securing Singapore’s leadership in AI Innovation

Meanwhile, domestic competition in China has intensified, compressing profit margins and pushing decision-makers to look abroad. Southeast Asia’s geographical proximity and cultural affinity make it a natural first choice.

Policy dividends are another driver. The Regional Comprehensive Economic Partnership (RCEP) has lowered tariffs, streamlined customs procedures, and strengthened supply-chain connectivity across the region. Chinese firms are also aligning with national growth strategies; for instance, Singapore’s push to host regional headquarters even as companies manufacture in neighbouring countries, or the digitalisation and green transformation agendas being rolled out across ASEAN.

Finally, leading Chinese companies are confident they can replicate domestic success abroad. Having served hundreds of millions of consumers at home, they bring valuable experience in business models, branding, and supply-chain management to Southeast Asia’s emerging but fragmented markets.

A changing cast of investors

Li has observed a notable shift in who is expanding. Two decades ago, the landscape was dominated by large state-owned enterprises and centrally linked conglomerates, concentrated in energy, natural resources, and large-scale infrastructure — capital-intensive projects tied to national strategic objectives.

Today, private companies with more flexible decision-making and faster execution are emerging as the new driving force. A growing number of mid-sized and smaller private firms — many technology-driven, innovation-focused, or consumer-manufacturing oriented — are becoming frontline players. They are building brands, localising products, and tapping into Southeast Asia’s rising middle class across sectors ranging from smart home appliances and intelligent furniture to higher-value-added consumer goods.

That said, manufacturing remains one of the largest pillars of Chinese overseas investment. New energy vehicle (NEV) manufacturers, battery and spare parts producers, and the broader green energy supply chain continue to account for a substantial share of activity, even as tech-enabled and consumer goods companies increasingly drive deal volume and diversification.

Offence, defence, or both?

Asked whether this expansion is driven by genuine growth ambitions or by companies routing around tariffs and geopolitical risk, Li says both motivations have coexisted in recent years, though the balance has shifted.

When tariffs on most Southeast Asian countries rose in 2025, alongside tightened enforcement on origin verification, Southeast Asia stopped being viewed merely as an intermediate transit point for channelling exports to global markets. Chinese companies are no longer just seeking cost reduction; they are strategically expanding production, building ties with local consumers, and integrating into local business ecosystems for sustainable, long-term growth.

Localisation over geopolitics

On concerns about being perceived as “too close” to Beijing, or caught in US-China dynamics, Li is clear: current expansion is driven by growth considerations rather than geopolitical ones. Southeast Asia is increasingly seen not as a low-cost assembly hub for exports, but as a core strategic pillar where Chinese companies can build locally rooted, resilient, consumer-focused businesses capable of thriving independently.

Also Read: Singapore outsmarts the world in AI–ranked No.1 global hub

This is reshaping how companies structure their regional entities. Decisions on investment vehicles, ownership arrangements, and brand development are increasingly guided by localisation strategies. Many companies favour joint ventures with local partners for better market access, while branding leans towards local consumer tastes, often while retaining certain distinctive Chinese characteristics.

From tax structuring to strategic advisory

KPMG China’s role has evolved alongside these shifts. For years, the firm has supported cross-border M&A for Chinese outbound investors, working with the KPMG global network to provide integrated advisory services spanning financial advisory, due diligence, valuation, tax structuring, post-investment integration, and ongoing accounting and tax support.

But as greenfield investment gradually overtakes M&A as the dominant mode of Chinese expansion, particularly in Southeast Asia, KPMG is seeing a surge in mandates to help clients build new operations from the ground up. This includes site selection, joint venture partner vetting, and facilitating communication with local authorities to secure approvals and incentives.

Spotting trouble early

When expansion goes wrong, the fallout typically emerges within a year or two, says Li. Common failure patterns include financial strain from over-investment and underperformance, compliance gaps stemming from regulatory missteps and legal disputes, and talent loss driven by cultural integration challenges.

Early warning signs include eroding trust and communication between shareholders and management, cost overruns, weaker-than-expected market response, notable employee turnover, especially among core management and key technical staff, and rising disagreements with local partners.

Taken together, Li notes, these signals underscore the importance of decision-makers consistently reviewing business assumptions and promptly adjusting strategy as they navigate their overseas expansion.

The post It’s not just tariffs: The real reason Chinese capital is flowing into ASEAN appeared first on e27.

Southeast Asia’s financial services dealmaking held steady in the first half of 2026, with 31 disclosed M&A deals matching 2025 levels, even as disclosed deal value fell to US$936M from US$1.6B, according to EY’s latest financial services M&A analysis.

Wealth and asset management emerged as the standout gainer: deal volume nearly tripled to eight from three, while disclosed value jumped to US$145M from just US$800,000 a year earlier, reflecting rising demand from the region’s growing affluent population across Singapore, Indonesia, Vietnam, Malaysia and Thailand.

Banking and capital markets, still the largest segment by value, cooled to 14 deals worth US$669M, down from 20 deals worth US$1.1B, as regulation and legacy-system complexity slow full-scale acquisitions.

Insurance recorded more deals (nine versus eight) but smaller total value (US$123M versus US$478M). Foreign acquirers grew more selective (five deals, down from seven), but committed larger sums, US$410M versus US$344M, signalling international appetite for Southeast Asian financial assets persists despite near-term caution. EY expects larger transactions to return in the second half of 2026 if financing conditions improve.

REGIONAL

Malaysia stays Grab’s top market with US$622M in H1 revenue: The ride-hailing and delivery giant generated more revenue from Malaysia than any other market in the first half of 2026, underscoring the country’s outsized role in Grab’s recovery story.

ZUS Coffee weighs IPO to raise at least US$245M: The Malaysian coffee chain is exploring a public listing that could value it significantly higher than its last private round, signalling renewed investor appetite for SEA consumer brands.

VinFast and Gowa Motor to open 30 EV showrooms in Indonesia: Vietnam’s EV manufacturer is forming a joint venture with Indonesia’s Gowa Motor to accelerate its regional expansion amid intensifying competition from Chinese rivals.

Vietnam and Singapore firms to digitise cross-border LC verification: Technology companies from both countries are partnering to automate letter of credit verification, targeting a process still dominated by paper and manual checks.

Malaysia’s sovereign AI bet: Local context is the new startup moat
Malaysia’s AI Nation 2030 plan wants the country to move from an AI buyer to a producer, betting that locally trained models fluent in Bahasa Melayu can outperform generic global platforms in high-value domestic use cases.

Malaysia maps an inclusive AI future for farmers and small shops
Malaysia’s National AI Action Plan aims to give 1.5 million MSMEs and smallholders scalable access to AI tools, targeting up to 70% less manual labour in palm oil operations through shared platforms and vetted datasets.

Deepgram sets up APAC base in Singapore for multilingual Voice AI
Deepgram is establishing its Asia Pacific headquarters in Singapore, backed by a strategic investment from EDBI, after seeing a 96% year-on-year jump in API requests across the region’s 20-plus markets.

---

INTERVIEWS & FEATURES

It’s not just tariffs: Why Chinese capital is flowing into ASEAN: KPMG’s Lisa Li tells e27 that Chinese firms are moving beyond tariff-dodging, drawn by Southeast Asia’s consumption potential and cheaper land and labour, with private firms now outpacing state-owned players.

Datakrew’s Hyundai pilot tackles the EV battery forecasting problem: Singapore’s Datakrew wrapped a year-long study with Hyundai CRADLE and GetGo, pulling 3.6B data points from 70 EVs to forecast battery health months in advance, within a claimed 3% prediction-error band.

Four lessons from GITEX Global: What Dubai’s AI playbook means for SEA: Sam Altman and UAE minister Omar Sultan Al Olama’s GITEX sessions revealed talent, not capital, as the real AI bottleneck, alongside energy supply emerging as AI’s next major constraint.

---

INTERNATIONAL

Anthropic’s annualised revenue surges to US$6.5B: The AI safety startup has seen explosive commercial uptake, with revenue growing rapidly as enterprise demand for Claude accelerates globally, putting pressure on OpenAI’s market dominance.

Uber and Pony.ai plan 2,000 robotaxis across Europe: The ride-hailing giant is expanding its autonomous vehicle push into European markets with Chinese AV firm Pony.ai, a deal that could reshape its long-term driver model.

SoftBank pours US$200M into construction robotics firm Gravis: The Japanese conglomerate’s latest bet targets construction automation, a sector it sees as ripe for disruption given chronic global labour shortages.

Groq valued at US$3.5B after fresh funding round: The AI chip inference startup closed a new round following its Nvidia deal, with investors betting its speed advantage over GPU-based rivals will hold as LLM demand scales.

Wispr raises US$280M at US$2B valuation beyond dictation: The voice AI startup is pushing into broader agentic use cases after its rapid growth in speech-to-text, backed by significant new capital at a unicorn valuation.

SpaceX officially closes its Cursor acquisition: The move confirms Elon Musk’s aerospace company is expanding into developer tools, with Cursor’s AI coding platform now folded into SpaceX’s growing technology portfolio.

Uber adds Zipline drones to its Eats delivery network: Zipline’s autonomous drones will fulfil food delivery orders through Uber Eats, marking a meaningful step toward commercial drone logistics at scale.

YouTube to count views from the first second of playback: The platform’s updated view-counting policy will affect creator metrics and ad measurement, with knock-on implications for how brands allocate digital spend.

Meta faces trial over social media addiction claims: Facebook and Instagram are under legal scrutiny in a landmark US trial that could set precedent for platform liability over user harm and algorithmic design.

AI automation startup Relay shuts down, staff joins Google Chrome: Despite early traction, Relay could not survive in an increasingly crowded agentic AI market, with its team acqui-hired into Google’s browser division.

---

SEMICONDUCTOR

Nvidia invests US$1.5B in SoftBank’s data centre unit behind OpenAI project: The chipmaker’s stake in SoftBank’s data centre developer deepens its position across the AI infrastructure stack and signals growing alignment between two of the industry’s most influential players.

NXP breaks ground on expanded Malaysia factory at 900,000 sq ft: The Dutch chipmaker’s facility expansion signals sustained foreign investment in Malaysia’s semiconductor manufacturing base, reinforcing the country’s role in global chip supply chains.

AI demand lifts Malaysia’s chip sector, but not every player wins
HSBC research shows Malaysia, Singapore and Vietnam are primary beneficiaries of the AI infrastructure boom, though companies without direct AI exposure face rising memory-chip costs squeezing margins instead.

CYBERSECURITY

Apple users hit by spyware alerts in unprecedented numbers: Security investigators report a surge in spyware notifications sent to Apple users globally, raising urgent questions about the scale and origin of the targeting campaign.

AI

Agentic AI adoption doubles to 51% among Singapore firms: A ServiceNow survey found that more than half of Singapore businesses now deploy agentic AI, the sharpest year-on-year jump recorded, outpacing adoption rates in most comparable economies.

Alipay launches agentic commerce platform for Chinese merchants: The payments giant is giving merchants AI-powered tools to automate customer interactions and transactions, a move that could influence how SEA super-apps evolve their merchant offerings.

Anthropic CEO frames AI backlash as a trust crisis: Dario Amodei argues that public resistance to AI is not about the technology itself but about institutional credibility, calling on the industry to prioritise transparency over capability announcements.

Zuckerberg’s AI vision fails to convince sceptics: Analysts and observers push back on Meta’s AI roadmap, questioning whether its open-source strategy and consumer AI products can generate the returns the market expects.

---

THOUGHT LEADERSHIP

SEA solved distribution: Now fintech must scale on the balance sheet: Fathhi Mohamed argues Southeast Asia’s public payment rails have turned fintech into a utility business, where cheap, sticky deposits now matter more than another million app downloads.

Not every cheque keeps every door open for SEA founders: Surabhi Pandey writes that a funding round now carries a geopolitical footprint, urging founders to run due diligence on investors the way investors scrutinise them, to protect long-term strategic optionality.

AI won’t just automate tasks, it will repackage responsibilities: Daniel Tan argues the deeper shift isn’t task automation but the redesign of recurring responsibilities into reviewable, delegable systems, changing how individual contributors must think about ownership.

Your AI isn’t producing bad creative, your brief is: Aleks Farseev cites WARC research showing 88% of marketers produce more creative with AI, yet only 45% see a genuine quality lift, blaming stale demographic-only briefing.

Why Southeast Asia doesn’t need to pick a side in the AI race: Jan Alvin Pabellon argues the region’s advantage lies in connecting rival tech ecosystems rather than choosing between them, building trust infrastructure instead of competing on model scale.

AI is not the advantage, build what competitors cannot copy: Christopher Jackson urges SEA SMEs to protect know-how through trade secrets or patents rather than relying on AI-generated polish, which rivals can replicate just as quickly.

Strategic chokepoints: Designing leverage without owning everything: Niharika Ray argues durable market power comes from sitting at points where uncertainty must be resolved — trust, compliance, settlement — rather than from owning the entire value chain.

Social intrapreneurs can change the world too: Dr Erwin Chan writes that entrepreneurial change inside organisations doesn’t require founding a startup, just the discipline to pilot a scoped idea using existing budget and headcount.

The US$46,300 question: How low can Bitcoin go before buyers return: Anndy Lian notes Bitcoin’s Coinbase premium has stayed negative for 90 days as AI infrastructure spending diverts capital away from crypto, with a bear pennant pointing toward US$46,300.

Despite the rally, the CMC Fear and Greed Index stays at 39: Anndy Lian writes that Bitcoin’s 2.23% surge to US$64,300 and a cooling Fed rate outlook haven’t shifted sentiment, with the index still signalling market fear despite bullish technicals.

The post Ecosystem Roundup: SEA financial services deals hold firm despite sharp drop in value appeared first on e27.

The conversation around artificial intelligence has become strangely polarised. One camp believes AI will replace millions of jobs. Another believes it’s the greatest productivity revolution in modern business.

I believe both camps are asking the wrong question.

After spending the last few years working alongside CEOs, Managing Directors and executive teams across financial services, technology and healthcare, I’ve come to a different conclusion.

AI won’t replace leaders. It will expose the ones who were never truly leading in the first place.

During a recent executive roundtable with multinational organisations, one question dominated the discussion.

“Did we spend too much, too fast, on AI?”

Some organisations had invested between US$200,000 and US$500,000 implementing AI platforms to gain first-mover advantage since late 2024. Yet despite sophisticated technology, many struggled to demonstrate measurable returns, till now.

The technology wasn’t broken. There just wasn’t a clear plan on what business outcomes AI was “supposed” to contribute to. The leadership operating system was faulty.

Technology moves faster than human transformation

Throughout history, every major technological leap has promised greater productivity. The Industrial Revolution automated manual labour. The internet democratised information. Cloud computing transformed collaboration.

AI is different because it is beginning to automate thinking itself. Yet while technology evolves exponentially, leadership capability often evolves incrementally, and in some cases, stays the same.

Many organisations have upgraded their technology stack without upgrading the way their leaders think, communicate and create trust for its implementation. This has created resistance to the immersive use of AI.

Now we have a dangerous gap. AI can generate reports. It cannot generate motivation. AI can analyse data. It cannot build psychological safety. AI can recommend decisions. It cannot inspire people to believe in them.

These have always been human responsibilities. So now, they have become competitive advantages. Imagine, would you have thought deep human connection to be a corporate advantage?

Also Read: The system behind the smile: How to make volunteer efforts sustainable

For years, organisations rewarded leaders primarily for technical expertise, operational efficiency and execution. Those capabilities remain important, but AI is rapidly commoditising technical knowledge.

When everyone has access to intelligence, leadership strength becomes the differentiator.

Leadership strength will be the ability to remain calm amid uncertainty. To think strategically when information is overwhelming. To communicate with clarity when ambiguity increases. To regulate emotion before making critical decisions. To create cultures where innovation feels safe rather than threatening.

This is why I have observed a consistent pattern. You cannot install AI on an outdated leadership operating system.

The organisations succeeding with AI are not necessarily those with the biggest technology budgets.

They are the ones who have leaders capable of helping people navigate uncertainty without losing trust, tying everything to a business outcome and not “AI for show”.

This is the biggest misconception about AI adoption, it is that implementation is primarily a technology project.

It isn’t, it’s a leadership transformation project. Employees rarely resist technology. They resist confusion. They resist poor communication. They resist feeling excluded from decisions that affect their future.

The highest-performing organisations don’t simply deploy AI. They create environments where people understand why change matters, how they contribute, and where psychological safety allows experimentation without fear.

Technology scales processes. Leadership scales people. The organisations that master both will outperform those investing exclusively in one.

From organisations competing through efficiency, today, they compete through adaptability. Tomorrow, they’ll compete through humanity.

Also Read: Why Singapore’s AI finance race is now about data, not models

As AI continues to level the playing field, qualities once considered “soft” will become remarkably hard to replicate.

• Empathy.
• Compassion.
• Presence.
• Influence.
• Resilience.
• Authenticity.

These are no longer leadership buzzwords but strategic assets.

Ironically, the more advanced AI becomes, the more valuable deeply human connection becomes.

Our most important question isn’t: “How quickly can we implement AI?”

It’s: “Are our leaders ready to lead in an AI-powered world?”

Because organisations don’t transform through technology. They transform through people who know how to lead technology. AI will undoubtedly reshape every industry, but it won’t replace exceptional leaders.

It will simply reveal who has been relying on authority, expertise or hierarchy instead of genuine influence.

The future belongs to organisations that invest as intentionally in leadership as they do in digital capability. The winners of the AI era won’t necessarily have the most expensive set ups, they’ll have the strongest leaders.

—

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 AI won’t replace leaders, but it will expose weak leadership appeared first on e27.

Startup Bangladesh, the government-backed venture capital and fund management company under the ICT Division, has begun operational activities for the Bangladesh Fund of Funds, a new initiative designed to channel state capital through professional venture capital fund managers rather than only making direct investments into startups.

The fund has an initial size of BDT 400 crore (~US$33 million). Its launch through a Request for Expression of Interest was announced at an event in Dhaka attended by government officials, development partners, venture capital and private equity firms, startup founders, and investors.

Also Read: 20 game-changing startups driving Bangladesh’s innovation wave

For Bangladesh’s startup ecosystem, the structure matters as much as the amount. A fund-of-funds does not typically invest directly into companies. Instead, it backs venture capital funds, which then invest in startups. If executed well, this can help create more professional fund managers, improve investment discipline, and bring in additional private and institutional capital alongside public money.

That is the larger bet behind the Bangladesh Fund of Funds. The government wants each unit of public capital to attract more local, international, and development finance into the country’s startup market, where foreign investors have historically supplied the bulk of funding.

Over the past decade, Bangladeshi startups have reportedly raised around US$1.2 billion. But local investors accounted for only about 7 per cent of the capital deployed. That gap has long been a weakness for the ecosystem: founders often depend on foreign funds for growth rounds, while domestic pools of risk capital remain thin.

The new vehicle is meant to address that bottleneck by supporting selected fund managers who can deploy capital across a broader base of startups.

A policy shift from direct support to market-building

The launch comes as Bangladesh places greater political weight on startups and entrepreneurship as part of its economic development agenda. The government’s 2026 election manifesto emphasised startup growth, job creation, innovation, and the development of a technology-led economy.

In the current fiscal year, the government has allocated BDT 500 crore (roughly US$41 million) for startup development. It has also introduced tax and VAT incentives, including a zero per cent turnover tax, to lower the burden on young companies.

These measures come at a time when startup funding across much of Asia has become more selective. After the liquidity boom of 2020 and 2021, venture investors have shifted towards profitability, stronger governance, and clearer paths to scale. In Southeast Asia, this has pushed founders to raise smaller, more disciplined rounds and forced governments to think beyond grants and ad hoc startup programmes.

Bangladesh appears to be taking a similar route by trying to build financial infrastructure around its startup economy. The fund-of-funds model is already familiar in more mature markets, including Singapore, where public capital has often been used to crowd in private investors and support emerging fund managers. For Bangladesh, the challenge will be to adapt that model to a younger market where fund management capacity, exit pathways, and institutional investor participation are still developing.

Fakir Mahbub Anam, Minister for Posts, Telecommunications and Information Technology, described the Bangladesh Fund of Funds as a major platform for connecting entrepreneurs with capital and networks.

Also Read: Bangladesh’s startup ecosystem is entering a new phase of investability

“It will help connect promising Bangladeshi entrepreneurs with the capital, expertise, and global networks they need to grow,” he said at the event. “Through this initiative, we want to build a stronger pathway for innovation-led enterprises to create employment, attract investment, and contribute to Bangladesh’s future economy.”

Why fund managers matter

One of the less visible problems in emerging startup ecosystems is not only the lack of money, but the lack of experienced intermediaries to allocate it. Venture capital depends heavily on judgement: which founders to back, how to price risk, when to support follow-on rounds, and how to help companies navigate hiring, governance, expansion, and exits.

By investing through professional fund managers, Startup Bangladesh is signalling that the ecosystem needs more than a state chequebook. It needs investors who can repeatedly source deals, build portfolios, work with founders, and attract co-investors.

Nurul Hai, Managing Director and CEO of Startup Bangladesh Limited, said the initiative is intended to strengthen the deeper plumbing of the market.

“The Bangladesh Fund of Funds is not just about providing capital,” he said. “We want public capital to unlock much larger pools of private and international investment, strengthen professional fund managers and give high-potential Bangladeshi startups a clearer path to scale.”

The proposed structure includes fund-manager selection, co-investment mechanisms, and a sidecar facility, according to the presentation made at the event. Sidecar facilities are typically used to invest alongside a main fund or syndicate, allowing additional capital to follow selected opportunities without changing the core fund structure.

Japan International Cooperation Agency representative Morikawa Yuko said the fund could help deepen Bangladesh’s venture market by attracting institutional and foreign investment and bringing global VC firms into the ecosystem.

That external validation could prove important. Across Southeast Asia, development finance institutions, government-linked funds, and multilateral agencies have played a key role in supporting early venture ecosystems, especially where domestic pension funds, insurers, and family offices are still cautious about the asset class.

Bangladesh’s regional moment

Bangladesh is not usually grouped with Southeast Asia in a strict geographic sense, but its startup trajectory increasingly overlaps with the region’s. Its large young population, rising digital adoption, growing mobile payments activity, and dense urban consumer markets resemble the conditions that helped produce major tech companies in Indonesia, Vietnam, and the Philippines.

Yet Bangladesh has lagged behind those markets in venture depth. Indonesia has produced multiple unicorns and a relatively large local VC base. Vietnam has drawn strong interest from regional funds as a manufacturing and digital economy story. The Philippines has benefited from fintech and digital services growth, despite funding volatility. Bangladesh, by contrast, has produced notable companies in fintech, logistics, commerce, education, and health, but the capital stack around them remains less developed.

That makes the Bangladesh Fund of Funds both an opportunity and a test. If it backs credible fund managers, applies transparent selection criteria, and avoids political allocation of capital, it could help create a more durable venture market. If it becomes another top-down financing scheme without independent investment judgement, its impact may be limited.

The timing is also important. Regional investors are more cautious today, but they are still looking for underpenetrated markets with large domestic demand. Bangladesh, with a population of more than 170 million, remains one of Asia’s largest consumer markets. For startups, the question is whether that demographic scale can translate into venture-scale businesses.

Also Read: PulseTech delivers Startup Bangladesh’s first multi-fold return after revenue surge

The government’s role will be to reduce friction without crowding out private capital. That means supporting fund managers, improving tax clarity, encouraging exits, and giving institutional investors enough confidence to participate.

For now, the Bangladesh Fund of Funds marks a shift in ambition. Rather than backing individual startups one by one, the government is attempting to build a financing layer that can outlast a single budget cycle. Whether it succeeds will depend less on the announcement and more on who gets selected, how capital is governed, and whether private investors decide Bangladesh is ready for a larger seat at the regional startup table.

The post Bangladesh launches US$33M fund-of-funds to deepen startup capital pool appeared first on e27.

Bitcoin currently faces a highly complex market environment that puzzles many seasoned investors. The premier digital asset struggles to break above the US$70,000 price level last observed in May. A closer look at market indicators reveals deep underlying weakness despite favourable macroeconomic conditions.

The Coinbase Bitcoin Premium Index calculates the price difference between Bitcoin on Coinbase and Binance to estimate buying or selling pressure from the United States market. This specific index has stayed negative for 90 consecutive days. At the time of writing, the index stood at -0.1066 per cent.

A negative reading indicates that the asset trades at a lower price on Coinbase than on Binance. This persistent discount highlights a profound lack of domestic buying interest. This prolonged negative premium is a glaring warning sign for the broader cryptocurrency sector. Smart money clearly anticipates further downside risk and refuses to accumulate more digital assets at current valuations.

This steep decline occurred while broader financial markets celebrated new record highs. The Relative Strength Index remained largely below the neutral level, reflecting prevailing bearish sentiment. Bollinger Bands further supported the volatility that prevented the price from hitting a high bullish threshold. Even massive accumulation by large holders failed to reverse the downward trend.

Whale wallets bought 54,000 more coins since mid-June, but the price action ignored this aggressive accumulation. Buy-side support below the current price continues to erode rapidly. A significant concentration of buy orders below the market existed earlier, especially in June. This created a solid floor because buyers were prepared to absorb selling pressure if the asset dropped toward those levels.

Market participants have now removed or shifted many of those bids lower, leaving fewer orders directly beneath the price. The market liquidity buffer has weakened significantly with less buy-side support to cushion further declines. I consider this lack of underlying bid depth a major structural vulnerability. Order book dynamics clearly show that large players are stepping away from defending current valuation levels.

Also Read: Pokemon cards gained 22.8% while Bitcoin lost 20.7% and that gap should worry every investor

The digital currency had nearly everything going its way this week but remains on track to finish roughly three per cent lower. This divergence strikes a particularly discordant note, given the asset’s reputation as a high-beta proxy for technology stocks. Wall Street pushed to fresh record highs as inflation cools and traders dial back expectations for a Federal Reserve rate hike in September.

These conditions normally favour speculative assets. The digital currency fell from around US$65,000 on Monday to US$62,470 by Friday. The tech-heavy Nasdaq 100 closed the week approximately one per cent higher during the exact same period.

Last week produced a clean dovish signal, combining cooler inflation with a weakening labour market, as reflected in favourable producer price index and jobless claims data. This refusal to follow traditional equities is deeply concerning for momentum traders. This distinct decoupling suggests that internal market mechanics currently overpower external macroeconomic stimuli. The asset faces its own distinct demand problem, setting it apart from the broader stock market rally.

Michael Saylor serves as the executive chairman at Strategy, which holds the record as the largest public company holding this asset. He offered the clearest explanation for this divergence earlier this month. Saylor noted that an enormous amount of capital is currently flowing into artificial intelligence infrastructure. Companies such as Alphabet, Meta, and SpaceX represent the largest near-term headwinds for the digital currency.

The premier cryptocurrency and artificial intelligence currently compete for the exact same speculative and institutional capital. Artificial intelligence is winning this battle for investor attention right now. This massive capital rotation explains why the digital currency refuses to participate in the broader equity rally. I believe this technological distraction will continue suppressing digital asset prices until the artificial intelligence hype cycle naturally cools down.

Institutional investors simply prefer the tangible revenue growth of technology giants over the speculative store-of-value proposition during uncertain economic times. This sector rotation severely limits the liquidity available to alternative assets seeking robust capital inflows. Wall Street allocates billions to data centres rather than decentralised ledger networks.

Also Read: Is the US$63,750 line the only thing standing between Bitcoin and US$62,000?

Exchange-traded funds further illustrate this lack of institutional enthusiasm. United States spot exchange-traded funds recorded US$5.48 billion in net outflows in 2026. These funds have only recovered US$459.6 million so far in August, as of August 14. This massive capital exodus confirms that large funds are reducing their exposure.

The digital currency formed a smaller bear pennant around US$60,000 to US$65,000 since the June selloff. This formation represents another bearish continuation pattern that technical analysts monitor closely. A decisive break below the rising support of this pennant could accelerate the existing flag breakdown. The measured move points toward approximately US$46,300.

That calculation puts the broader downside target zone at roughly US$45,000 to US$52,000. I expect the market to test these lower support levels before finding any meaningful long-term stability. Traders must respect these technical breakdown signals and adjust their risk management strategies to protect their portfolios from sudden drawdowns. Chart patterns rarely lie, and this specific setup screams further downside action for anyone paying close attention.

The broader economic backdrop adds another layer of complexity to this situation. The United States national debt currently nears US$40T. This massive fiscal burden forces the government to issue more bonds, which drains liquidity from the financial system.

I argue that this expanding debt ceiling inherently restricts the amount of excess capital available for highly speculative assets. The combination of massive artificial intelligence investments and soaring national debt creates a perfect storm that suppresses digital asset valuations. Investors must recognise that the digital currency no longer moves in lockstep with traditional risk assets.

Market participants should prepare for increased volatility and potentially lower prices in the coming weeks. Prudent traders will likely hedge their portfolios against these impending macroeconomic shocks. The digital asset must overcome these significant structural headwinds before it can resume its historical upward trajectory. Careful observation of order book depth will provide the next major clue. Global liquidity constraints will dictate the next major move.

—

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 post The US$46,300 question: How low can Bitcoin go before buyers return appeared first on e27.

Malaysia is emerging as one of Southeast Asia’s clearest winners from the artificial intelligence infrastructure boom, but the benefits are not spreading evenly across its semiconductor industry, according to HSBC Global Investment Research.

In a report released last week, the research house said Malaysia, alongside Singapore and Vietnam, stands out as a primary regional beneficiary of the AI-driven technology upcycle. The evidence is visible in the country’s surging chip exports and its deepening electronics trade flows with the United States, mainland China, Taiwan and neighbouring Singapore.

Also Read: Chips, corruption, and credibility: Malaysia’s semiconductor gamble faces a trust test

Those flows matter because Malaysia sits at a critical point in the global chip supply chain. Fabricated chips are often shipped into the country for assembly, testing and packaging, or ATP, before moving on to device makers, cloud infrastructure providers and end customers. This back-end role has made Malaysia indispensable to global semiconductor production, even if the highest-margin parts of the industry still sit elsewhere.

But HSBC’s central point is more nuanced than a simple “AI boom lifts all boats” story. Malaysia is gaining from the global race to build data centres, train large AI models and secure computing capacity, yet the upside is concentrated among companies with direct exposure to AI-related demand.

A boom with uneven rewards

Firms linked to AI infrastructure, including selected chipmakers, advanced packaging providers and data centre operators, are seeing stronger demand as hyperscalers and technology companies continue to spend heavily on computing power.

The picture is less straightforward for companies tied to traditional consumer electronics. For them, the same AI boom can become a cost problem. Strong demand for memory chips, for instance, can push up input prices and squeeze manufacturers that rely on those components but do not directly benefit from AI-related orders. In some cases, higher memory costs could even slow production.

This split is important for Southeast Asia. The region’s electronics sector is often discussed as a broad beneficiary of supply-chain diversification and AI demand. In practice, exposure varies widely by product, customer base and position in the value chain. Malaysia’s semiconductor sector is large, but not every participant is equally plugged into the most profitable parts of the AI cycle.

HSBC said broader operating conditions remain resilient. The global electronics Purchasing Managers’ Index eased to 55.3 in July from 55.7 in June, while the Asia electronics PMI slipped to 54.7 from 55. Both remain firmly above the 50 mark that separates expansion from contraction, and are still ahead of their 12-month averages of 52.5 and 53.6, respectively.

That suggests the cycle remains healthy, even as pressure builds in specific parts of the supply chain. HSBC noted that input and output prices eased slightly in July and supplier delivery times improved marginally, but price pressures remain elevated and order backlogs are still growing.

The helium risk

A second concern is supply risk, particularly around key inputs used in chipmaking. HSBC highlighted lingering uncertainty linked to the Middle East conflict and its potential effect on critical materials such as helium.

Helium is used in parts of semiconductor manufacturing, especially in front-end wafer fabrication, where chips are created on silicon wafers. Any disruption to supply can therefore create complications for countries trying to expand fabrication capacity.

Malaysia is partly insulated because its largest semiconductor strength remains ATP, which depends less on helium-intensive processes and more on nitrogen. The country also has substantial domestic nitrogen production. Still, the risk is not irrelevant.

“Fortunately for Malaysia, ATP relies less on helium-intensive processes and more on nitrogen, for which the country has substantial domestic production,” HSBC said. “However, Malaysia’s wafer fabs do rely on helium, which means supply management still matters.”

Also Read: Malaysia’s GreatAsic raises US$6.9m to pivot nation from chip assembly to indigenous design

That distinction captures Malaysia’s current position well. Its dominance in back-end activities gives it resilience, but its ambition to move into more advanced front-end manufacturing exposes it to a different set of operational and geopolitical risks.

Moving up the stack

Malaysia’s long-term challenge is not simply to attract more semiconductor investment, but to capture more sophisticated parts of the industry. The government’s National Semiconductor Strategy, announced in 2024, commits US$6.12 billion and targets the training of 60,000 highly skilled local semiconductor engineers by 2030.

The strategy reflects a broader regional ambition. Southeast Asian economies want to move beyond being assembly bases and become deeper technology hubs. For Malaysia, that means expanding advanced packaging, building more front-end manufacturing capability and nurturing chip design talent.

The difficulty is that foundries are among the most capital-intensive industrial assets in the world. They require vast financing, reliable energy and water supplies, specialised infrastructure, and long-term customer commitments. These hurdles can be addressed over time with incentives and execution, but talent is harder to manufacture quickly.

HSBC flagged engineer retention as a key weakness. Average engineering wages in Malaysia’s manufacturing sector trail those in several Asian competitors, creating a clear risk of talent outflows, particularly to neighbouring Singapore, where pay is significantly higher.

Matching wages with wealthier economies will be difficult. But Malaysia can narrow the gap through targeted grants, tax incentives and schemes that improve the overall value proposition for skilled roles. It can also use targeted immigration policies to ease the talent constraint.

For founders, investors and operators in the region, this is the less glamorous but more decisive part of the semiconductor story. Capital announcements make headlines, but execution depends on whether countries can build and keep enough engineers, technicians and managers to run complex industrial ecosystems.

Malaysia’s geopolitical opening

One advantage Malaysia does have is geopolitical positioning. HSBC said the country’s ability to maintain constructive ties with both the US and China, while complementing Singapore’s more constrained land and resource base, could help it attract diversified foreign investment.

This matters as multinational chipmakers reassess location risk. Taiwan remains central to global semiconductors, South Korea is a memory powerhouse, and mainland China continues to invest heavily in self-sufficiency. But geopolitical tensions around major Asian chip hubs are forcing companies to think harder about redundancy and resilience.

Malaysia’s neutral reputation could therefore become a more valuable asset. Between January 2024 and March 2026, its semiconductor sector secured about US$22.52 billion in approved investments, including roughly US$20.31 billion in foreign direct investment.

The country’s roots in semiconductors go back more than five decades. Intel opened a chip assembly plant in Penang in 1972, shortly after Singapore entered the industry. Since then, Malaysia has grown into a major back-end semiconductor hub, accounting for 13 per cent of the global ATP market.

It is also, together with Singapore, one of only two ASEAN economies with front-end fabrication capabilities. Malaysia has particular strength in automotive power semiconductors, supported by German chipmaker Infineon’s manufacturing footprint in the country.

Also Read: Malaysia’s chip suppliers face rising pressure to prove cyber resilience

The scale is already substantial. Malaysia exported nearly US$110 billion of semiconductors in 2025, equal to around 23 per cent of gross domestic product.

The AI boom gives Malaysia a powerful tailwind. But HSBC’s report suggests the next phase will be harder than riding export momentum. The country must manage supply risks, avoid a two-speed industry, and solve the talent problem if it wants to move from being a crucial assembly hub to a higher-value semiconductor power.

The post AI demand lifts Malaysia’s chip sector, but not every player wins appeared first on e27.

Southeast Asia’s financial services dealmaking held its ground in the first half of 2026, even as the total value of transactions dropped sharply, suggesting that buyers remain active but more selective in a market still shaped by high uncertainty.

According to EY’s latest financial services M&A analysis, the region recorded 31 publicly disclosed mergers and acquisitions in the first six months of the year, unchanged from the same period in 2025. But disclosed deal value fell to US$936 million from US$1.6 billion a year earlier.

Also Read: ‘M&A process in SEA is stuck in the dark age’: say match.asia co-founders

The headline number points to a market that has not frozen, but has become more cautious. Rather than the large, transformative deals that dominated parts of the previous cycle, Southeast Asia’s financial services activity in early 2026 was led by smaller and mid-sized transactions. For banks, insurers, fintech investors and asset managers, the focus appears to be on assets that can fill specific strategic gaps rather than aggressive expansion at any price.

“The stability in deal activity across Southeast Asia’s financial services sector reflected a market that remains engaged despite economic and geopolitical volatility,” said Stuart Last, EY-Parthenon Partner, Financial Services, Ernst & Young Solutions LLP.

He added that the fall in disclosed deal value suggests investors are pursuing opportunities with a clear strategic rationale, rather than chasing scale for its own sake.

Banks slow, wealth platforms gain ground

The split across sub-sectors shows how investor attention is shifting within the region’s financial services market.

Banking and capital markets remained the largest contributor by value, but activity declined. Deal volume in the segment fell to 14 from 20 a year earlier, while disclosed value dropped to US$669 million from US$1.1 billion.

That decline is not surprising. Banking deals in Southeast Asia are often shaped by regulation, ownership limits and the complexity of integrating legacy systems. While the region’s banks are still under pressure to digitise, improve cost efficiency and compete with fintech players, full-scale acquisitions can be difficult to execute, especially when interest rates, credit risk and capital requirements remain in focus.

Insurance moved in the opposite direction by deal count. The sector recorded nine deals in the first half of 2026, up from eight a year earlier. But disclosed deal value fell to US$123 million from US$478 million, indicating that activity was concentrated in smaller assets.

The more striking change came from wealth and asset management. Deal volume rose to eight from three, while disclosed value jumped to US$145 million from just US$800,000 in the first half of 2025.

That rise reflects one of Southeast Asia’s most persistent financial services themes: the growth of the affluent and mass-affluent population. As income levels rise in markets such as Singapore, Indonesia, Vietnam, Malaysia and Thailand, more consumers are seeking investment products, retirement planning tools and advisory services beyond traditional savings accounts.

For acquirers, wealth platforms can offer access to sticky customer relationships, fee-based revenue and digital distribution channels. In a region where financial literacy and investment participation are still uneven, firms that can combine trust, technology and local market access are becoming more attractive targets.

Last said the sharp rise in wealth and asset management deal value points to growing investor interest in platforms and capabilities that can capture demand from the region’s expanding affluent population.

Foreign buyers remain interested in Southeast Asia

EY’s data also suggests that international appetite for Southeast Asian financial services assets has not disappeared.

The number of non-Southeast Asian firms acquiring targets in the region fell to five in the first half of 2026 from seven a year earlier. However, the total disclosed value of these deals rose to US$410 million from US$344 million.

Also Read: How M&A can supercharge your startup’s success

That means fewer foreign acquirers were active, but those that did move were willing to commit larger sums. This matters because Southeast Asia continues to be viewed as a long-term growth market despite near-term volatility. The region has a young population, rising digital adoption, a large underbanked base in several markets, and increasing demand for credit, insurance and wealth products.

At the same time, operating across Southeast Asia is rarely straightforward. The region is not a single market. Financial services firms must navigate different regulators, licensing regimes, consumer behaviours, languages and levels of digital infrastructure. That complexity can slow dealmaking, but it can also make established local platforms more valuable.

EY expects larger transactions to return in the second half of 2026 if financing conditions improve and more scaled assets become available.

Global deal count rises, but megadeals thin out

The Southeast Asian pattern mirrors a broader global trend: more deals, but less value.

Globally, banks, insurers and asset managers publicly disclosed 1,137 financial services deals in the first half of 2026, up 3 per cent from 1,101 a year earlier. Yet total disclosed deal value fell to US$134.5 billion from US$191.3 billion.

The drop was largely driven by a thinner pipeline of megadeals. EY recorded 25 transactions above US$1 billion in the first half of 2026, representing 80 per cent of total deal value. That compares with 37 such deals in the first half of 2025 and 55 in the second half of 2025.

The concentration of value among the largest transactions remained high. The ten biggest global financial services deals accounted for US$78.7 billion, or 58 per cent of total value. The top 20 deals accounted for US$100.5 billion, or 75 per cent.

Omar Ali, EY Global Financial Services Leader, said financial services firms have adapted to heightened uncertainty as part of normal operating conditions. But he noted that unpredictability, slower global growth, inflation and supply shocks continue to affect deal value.

“Despite the number of transactions rising, deal value in the first half this year across the world’s major markets is down on 2025 levels, as significantly fewer transactions completed over the US$1 billion mark,” he said.

Asia and Oceania weaken, but cross-border interest grows

Across Asian and Oceanian markets, the first half of 2026 was softer than in Southeast Asia. Publicly disclosed financial services M&A fell 14 per cent to 147 deals from 170 a year earlier. Total disclosed value slipped to US$15.8 billion from US$17.8 billion.

Also Read: M&A in Asia: A strategic roadmap for venture builders

Banking and capital markets deal volume in the broader region declined to 77 from 87, though deal value rose to US$11.3 billion from US$6.4 billion. Insurance weakened more clearly, with volume falling to 31 from 41 and value dropping to US$2.1 billion from US$5 billion. Wealth and asset management also declined, with volume falling to 39 from 42 and value sliding to US$2.4 billion from US$6.5 billion.

However, foreign interest in Asian and Oceanian targets increased. Non-regional acquirers completed or announced 28 deals, up from 23, while disclosed value rose to US$1.9 billion from US$1.6 billion.

For Southeast Asia, the message is mixed but not gloomy. Dealmakers are not retreating from the region. They are becoming more disciplined, more sector-specific and more careful about valuation. The next phase of activity may depend less on whether buyers have appetite, and more on whether sellers are willing to meet the market.

The post Wealth management emerges bright spot in Southeast Asia financial services M&A appeared first on e27.

For years, Southeast Asia’s digital economy has grown on top of technologies built elsewhere. Cloud infrastructure, operating systems, search, social media, e-commerce tools and, more recently, large language models have largely come from the US and China. The region adapted quickly, but rarely controlled the deepest layers of the stack.

Artificial intelligence is forcing governments and founders to revisit that bargain.

Also Read: The unexpected ways AI is already changing Malaysia’s economy

Malaysia’s National AI Action Plan 2026-2030, also referred to as AI Nation 2030, places this question at the centre of its digital strategy: can the country move from being a buyer of AI tools to a producer of AI products, infrastructure and talent? The plan’s answer is built around sovereign AI — the ability to build, train and deploy AI using domestic infrastructure, data and workforce capabilities.

This is not just a policy slogan. For startups, it could become a practical competitive advantage. In a region as linguistically and culturally complex as Southeast Asia, models trained primarily on foreign datasets often miss local nuance. That gap matters in areas such as education, public services, healthcare, agriculture, financial inclusion and legal compliance, where context is not a nice-to-have but the product itself.

Local context as a moat

The strongest argument for sovereign AI is not that every country needs to rebuild OpenAI, Google or Anthropic from scratch. It is that global models, however powerful, are not always designed for local realities.

Malaysia’s plan recognises this through its Secure and Localised AI Ecosystem initiative, which calls for trusted local datasets and indigenous AI models. One priority is the development of large language models fluent in Bahasa Melayu and aligned with Malaysian cultural norms, including through collaboration with institutions such as Dewan Bahasa dan Pustaka.

For founders, this creates room to build where global platforms are weakest. A generic chatbot may perform adequately in English-language customer support, but it may stumble when handling Bahasa Melayu, Manglish, code-switching, dialects, religious sensitivities or public-sector terminology. In classrooms, government offices or rural advisory services, those mistakes can damage trust.

The same logic applies across Southeast Asia. Indonesia, Thailand, Vietnam and the Philippines all face similar challenges: large populations, diverse languages, uneven digital literacy and public services that need to work beyond metropolitan centres. A model that understands a population’s language, regulations and social context can outperform a larger but less grounded system in specific high-value use cases.

This is where local startups may find their wedge. They do not need to win the global foundation-model race. They can build domain-specific AI tools that combine global advances with local data, workflows and compliance requirements.

From AI adoption to AI production

Malaysia’s ambition is also economic. AI Nation 2030 aims to place the country among the top 10 in global AI indices, add an incremental 1.2 percentage points to GDP growth and create 300,000 new jobs by 2030.

Those are aggressive targets, but they reflect a broader shift in Southeast Asia’s thinking. The region’s internet economy has grown rapidly — Google, Temasek and Bain estimated it at US$263 billion in gross merchandise value in 2024 — yet much of the value still accrues to platform owners, cloud providers and chipmakers outside the region.

Also Read: How can Malaysia leverage AI for growth and not see it as a threat?

Malaysia wants a larger share of the value chain. Its plan uses public-private partnership “Impact Engines” to push local firms into higher-value AI segments, rather than leaving them as downstream users of foreign tools. One example is the AI Hub for the Manufacturing Ecosystem, which builds on Malaysia’s existing strength in semiconductors and electronics.

This is a sensible starting point. Malaysia is already part of the global chip supply chain, particularly in assembly, testing and packaging. Applying AI to improve yields, predict maintenance issues, optimise energy use and manage supply chains could help local manufacturers move up the ladder. It may also make the country more attractive to high-value foreign direct investment at a time when companies are diversifying supply chains across Asia.

For startups, the opportunity lies in the middle layer: AI applications for factories, logistics providers, farms, banks, schools and government agencies that need solutions adapted to local operations. These are not always glamorous markets, but they are often where durable revenue is built.

Data and compute decide who gets to build

Sovereign AI ultimately depends on two scarce inputs: data and compute.

Malaysia’s plan proposes an AI-ready data ecosystem that aggregates priority datasets across sectors and turns them into trusted, purpose-driven data products. A National Data Exchange would allow startups and institutions to discover and license datasets under clearer terms.

If executed well, this could address one of the biggest constraints for AI startups in the region. Many founders can access open-source models, but not the high-quality local datasets needed to make those models useful. Data is often fragmented across ministries, state agencies, corporates and legacy systems. Access can be slow, opaque or legally uncertain.

The compute side is just as important. Training and fine-tuning AI models requires expensive hardware, and reliance on foreign cloud providers can become a strategic vulnerability for governments handling sensitive data. Malaysia’s proposed National Supercomputing Centre and AI Sukuk financing mechanism are designed to expand access to high-performance compute for local innovators.

The key question will be implementation. If data access remains bureaucratic or compute is captured by large incumbents, startups will see little benefit. But if the infrastructure is affordable and fairly governed, it could reduce the entry barrier for Malaysian AI companies and research teams.

Trust as a market advantage

AI adoption will also depend on public confidence. Malaysia’s plan includes a governance framework aligned with the National Guidelines on AI Governance and Ethics, alongside an AI Trust Function to monitor and respond to risks. It also proposes a National AI Classification mechanism for “Made by Malaysia” AI products.

For startups, this may sound like more compliance work. In practice, clear rules can help. Regulated sectors such as finance, healthcare, education and public services will not adopt AI at scale without confidence that systems are safe, explainable and accountable.

A national classification framework could become a trust signal for buyers, particularly if Malaysia wants its AI products to travel across ASEAN. The region is still early in building interoperable AI governance. A credible Malaysian certification could help local companies sell into neighbouring markets that face similar concerns but may not have the same institutional capacity.

Inclusion will determine the outcome

Malaysia’s sovereign AI strategy is not only about building elite technology. It also targets micro, small and medium enterprises, which make up 84.4 per cent of businesses in the services sector. These firms often lack specialist talent and cannot afford bespoke AI systems. The AI for MSMEs initiative proposes modular, pre-vetted tools embedded into platforms they already use.

Agriculture is another test case. A scalable agristack using local data could support precision farming, weather prediction and crop advisory services. For a country seeking stronger food resilience, AI will only matter if it reaches farmers, cooperatives and small suppliers, not just large agribusinesses.

This inclusive angle is important for Southeast Asia. If AI primarily benefits large companies in capital cities, it may deepen existing divides. If it improves productivity for small businesses, schools, clinics and farms, it could become a broader development tool.

Also Read: AI demand lifts Malaysia’s chip sector, but not every player wins

The final piece is talent. Malaysia’s plan includes a Global Talent Network, Digital e-Residency Pass, AI Talent Concierge and large-scale reskilling efforts. These measures acknowledge a hard truth: sovereign AI cannot be built with infrastructure alone. The country needs engineers, product managers, policy specialists, domain experts and teachers who understand both technology and local problems.

Malaysia’s sovereign AI push is ambitious, and many parts remain dependent on execution. But its underlying bet is sound: in AI, local context is not a weakness. It may be the clearest moat Southeast Asian startups have.

The post Malaysia’s sovereign AI bet: Local context becomes the next startup moat appeared first on e27.

Volume went up. Quality didn’t. The bottleneck moved somewhere nobody is looking.

Every marketing team I speak to in Southeast Asia has the same story about generative AI: we make ten times more creative than we did eighteen months ago, and we’re not convinced any of it is better. That instinct is now measurable. New research from WARC with TikTok and LIONS Advisory, reported by PPC Land, surveyed 400 marketers and found that while 88 per cent are producing more creative since adopting AI, only 45 per cent see a meaningful lift in quality.

The interesting part isn’t the gap. It’s the reason for it. Two-thirds of those marketers said they brief generative models primarily with demographic data. Nearly six in ten of the same group said demographic segmentation no longer works. They are feeding the machine inputs they have already told researchers are broken. Only 17 per cent consistently brief with anything richer — community context, behavioural signal, actual audience tension.

The report’s framing is the sharpest line in it, and it deserves repeating in plain terms: this is not a technology gap. It is an intelligence gap.

The evaluation layer nobody costed

Here is what changed, and why it is easy to miss.

A brief has never really been a document. It has been the start of a conversation. It went to a strategist, who pushed back on the segment. It went to a creative director, who asked what the person actually feels at the moment of purchase. It went to an art director, who threw out the first three routes. By the time work reached a client, a thin brief had been quietly repaired four or five times by people whose job was, in part, to notice that it was thin.

Those stages were slow, and slowness was the point. They were also expensive, and so they were the first thing compression removed. When production timelines collapse from three weeks to three days, the repair layer goes with them. The brief no longer passes through a series of sceptical humans. It passes into a model, which is constitutionally incapable of scepticism about its own inputs and will produce forty confident variants of a bad idea as readily as forty of a good one.

So the weakness that used to be absorbed by the process now lands directly in the output, at volume, and with the polish of professional work. That is a much worse failure mode than the one it replaced. A bad brief used to produce visibly bad work that somebody caught. Now it produces plausible work that nobody catches, because it looks fine.

Also Read: Moving past the chatbox: The hidden risks of agentic AI and MCP in enterprise infrastructure

Demographics survive because they are already in the template

Why do teams keep briefing with data they don’t believe in? Not conviction — inertia.

Age brackets and income bands are already sitting in the planning deck. They are already in the media plan, the audience field, the campaign naming convention. They require no new work, no new tooling, no argument with anyone. Behavioural and community insight requires all four. Under deadline, the default wins every time, and the default is a demographic.

GWI made this point publicly last week in a rather good line — that age brackets are the laziest segment in marketing, and the differences inside a generation are larger than the differences between generations. They are right, and it is worth noticing how rare that argument is. Look across a week of published content from the marketing-technology category and a clean division appears. The audience research firms publish findings that stop at the statistic. The workflow and listening platforms publish features that start at the publish button. Almost nobody addresses the space between the two, which is precisely where the intelligence gap lives.

Why Southeast Asia feels this first

Because the compression here is more severe. A regional team in Singapore is routinely running six to eleven markets, in several languages, against budgets that would cover two markets in Europe. Local nuance is not a refinement; it is the entire job. And it is exactly the layer that demographic briefing flattens.

Feed a model “women 25–34, urban, middle income” and it will return something that could run in Jakarta, Manila or Kuala Lumpur and land properly in none of them. The output will be grammatical, on-brand and completely generic. Multiply that by eleven markets and a weekly cadence and you have built a very efficient machine for producing content nobody remembers.

Also Read: No fans, no fridges, just paint: ZERC’s founder on cracking SEA’s cooling crisis

Rebuilding the layer, cheaply

The teams pulling ahead have made one structural change: they treat the brief as the product. Not the deck, not the asset — the brief. Whoever controls the quality of the input now controls the quality of everything downstream.

That change is showing up in the tooling at both ends of the pipeline. At the front, platforms that decode live category data into audience tensions and evidence-backed briefs, so the input carries something observed rather than something assumed. At the back, a newer class of businesses built purely around execution — Touchigh, for instance, which helps Chinese cross-border sellers reach American buyers by automating both AI-search visibility and native English social content, and scores that content for predicted performance before it publishes rather than reporting on it after.

That last detail matters more than the category it sits in. A young execution company, serving SMEs at a few hundred dollars a month, has independently arrived at the same conclusion: the judgement call has to happen before the asset ships, or it doesn’t happen at all. Front end and back end are solving different halves of one problem — stopping weak inputs entering the system, and stopping good ones degrading on the way out.

What the operating numbers suggest is that the repair layer can, in fact, be rebuilt cheaply enough to survive a deadline. Pitching a UK hospitality group against considerably larger shops, the agency SAMY documented category planning collapsing from weeks to roughly thirteen minutes per brief, with predicted click-through accuracy running about three times better than human estimation. They won the retainer.

Read that carefully, because the speed is not really the story. The story is that the judgement nobody could afford is now something you can afford on every single brief — and a judgement that only happens when there’s time for it isn’t a standard, it’s a luxury.

The question worth sitting with

The uncomfortable version of the WARC finding is that most teams already know their inputs are wrong and ship on them anyway, because fixing the input costs more this week than shipping the output does.

That maths is changing. When evidence-grade audience intelligence takes minutes rather than weeks, the excuse for briefing on a demographic quietly disappears — and so does the defence when the work underperforms.

So: of the briefs your team wrote this quarter, how many could you trace back to something you actually observed about the audience, and how many were the template with a new date on top?

—

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 AI isn’t producing bad creative, your brief is appeared first on e27.