For decades, the architecture of international education ran in one direction. Asian families sent their children west, and the financial plumbing followed. Tuition flowed out of emerging markets into the bank accounts of universities in London, Boston, Sydney, and Toronto. The infrastructure was built around that single assumption.

It no longer holds. Students across Asia are increasingly choosing to study closer to home. Singapore is drawing record interest from South Asian applicants. Hong Kong’s universities, openly naming Singapore as a rival, are moving to lift their non-local enrolment ceiling to 50 per cent of local places. Malaysia, Vietnam, South Korea and Japan are all positioning themselves as destinations rather than only sources. None of this is a temporary post-pandemic blip. The pivot east is structural, not seasonal, and it is moving faster than the headlines.

Asia is no longer just a source of international students. It is rapidly becoming a destination for its own.

The payment rails carrying tuition between Asian families and Asian universities were not built for any of this.

The scale and the shift

The numbers are not small. India alone sends over a million students abroad each year. China has a similarly large outbound population, spread across more than 80 countries. Southeast Asia adds another 350,000-plus, making it one of the fastest-growing source regions globally. HolonIQ values the international education market at US$196 billion annually, projected to reach US$433 billion by 2030.

For most of these families, funding an overseas education is the largest cross-border transaction of their lives. Tuition, accommodation deposits and administrative fees get bundled into a handful of high-value transfers, executed under tight timelines and complicated regulation. Increasingly, both ends of that flow sit inside Asia. The infrastructure carrying it does not.

A system optimised for the wrong thing

The payment stage is where an offer-holder becomes an enrolled student. A failed transaction here is a lost enrolment. As a result, most institutions and payment providers optimise for a single outcome: whether the transaction goes through.

That is the wrong test. A payment going through is necessary; it is nowhere near sufficient. Conversion is the floor. It should not be mistaken for the ceiling.

Also Read: SEA’s SMEs aren’t lazy, but their payments infrastructure is

What families actually need is something harder to measure. They need to feel confident throughout. When a household is moving the largest sum of money it has ever moved, often for the first time, into a foreign banking system on a deadline tied to a visa, the absence of confidence is itself a failure mode, even if the money eventually arrives.

Asia’s payment paradox

The gap is particularly striking in Asia, home to some of the world’s most advanced domestic payment ecosystems. India’s Unified Payments Interface processed a record 21.6 billion transactions in December 2025 alone. Alipay and WeChat Pay have made cash close to optional across urban China. QR-based mobile wallets have rewired everyday commerce across Southeast Asia in under a decade.

Almost none of this translates across borders.

A Vietnamese family paying tuition to a campus in Singapore, or an Indian family wiring fees to Japan, is moving money between two digitally sophisticated payment economies, and still falling back on slow, opaque bank rails. Transfers route through multiple intermediaries. Foreign exchange costs are difficult to understand. Visibility into when funds will arrive is limited. The result is a fragmented experience layered on top of otherwise advanced financial systems.

Where friction becomes risk

For most families, this is not a transaction. It is a commitment, often the largest single transfer they will ever authorise, tied to a child’s future and a deadline they cannot move. And it is happening at a moment when households across the region are scrutinising every outgoing dollar.

The Chinese economy is in its longest stretch of consumer caution in two decades. Indian families are weighing rupee depreciation against rising overseas tuition. Across Southeast Asia, the post-pandemic squeeze has not fully lifted. In that context, an opaque foreign exchange margin is not a minor cost. An unexplained two-day delay is not a minor inconvenience. These are the moments where a family begins to wonder if they made the right choice. Some end worse: a missed enrolment deadline, or a switch to an informal channel that is faster but less safe.

Also Read: Digital payments: Adapting to a changing world

The institutional side wears the same friction differently. Tuition is high-value and compliance-sensitive. Every payment has to be reconciled against a student record, audited, and often chased through three or four intermediaries before it can be matched. Finance teams burn capacity on exception handling that legacy infrastructure was never designed to absorb at this scale.

A business-model problem dressed up as a technology one

Yes, capital controls and FX limits shape these flows. Mainland Chinese students operate under a US$50,000 annual personal quota. India’s remittance regime carries its own reporting obligations. Most emerging-market currencies come with constraints of some kind.

But these have always existed, and they do not explain the friction. The infrastructure to move money across borders, within these rules, already exists. What has been missing is any commercial reason to design it around the user, around a family in Surabaya wiring US$30,000 to Singapore, rather than around a global bank’s correspondent relationship in New York.

The challenge is less about whether payments can be completed and more about how they are experienced. Can the sender see where the money is at each stage? Are the fees and exchange rates legible? Does the confirmation arrive in time to matter? These are not edge cases. They are what trust is made of.

The metric no one is tracking

Payment success rates show up in every dashboard. Confidence does not. Yet confidence is what families remember.

A student who pays successfully but spends three days uncertain whether the money has arrived carries away a different impression of the institution than one who experiences clarity throughout. Across Asia, where decisions about where to study are shaped by agents, family WhatsApp groups and word of mouth more than by rankings or marketing, those impressions travel. Quickly, and across borders.

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

A market the incumbents were not built for

Cross-border education payments are a strange overlap: high-value, recurring, family-led, compliance-heavy, and increasingly intra-Asian. The legacy players solved a different version of this problem, one that ran from emerging markets to Western universities, in a world that no longer exists. The space is still shaped by their assumptions even as the flows move away from them.

Improving outcomes from here is not about adding more payment methods to the existing stack. It is about rethinking what visibility, trust and usability look like when both ends of the transaction sit inside Asia.

Raising the standard

As Asia’s role in global student mobility expands, the standard for the payment systems underneath it has to expand with it. A transaction that completes is no longer enough. The question worth asking is whether the system delivers confidence to the family on the other end.

A simple audit, for any institution or fintech reading this: at every stage of an inbound education payment, what does the family actually see, and what do they not?

—

The post Asia’s student boom is exposing a hidden weakness in global payments appeared first on e27.

For the past few years, the public image of artificial intelligence has been shaped almost entirely by the chatbot. People type questions; AI answers. People ask for summaries, emails, scripts, ideas; AI produces words.

This is genuinely powerful. But it is prologue.

The real future of AI is not conversation. It is action, and the distance between those two things is wider than most people appreciate.

LLMs are powerful, but they are mostly passive

Large Language Models are extraordinary instruments of language. They explain, translate, summarise, code, and communicate with a fluency that would have seemed miraculous a decade ago. But for all their sophistication, most LLMs are fundamentally passive.

They wait for instructions. They respond to prompts. They produce possibilities. They do not naturally perceive the physical world, test hypotheses in real environments, or improve through direct consequence the way humans and animals do.

Consider the difference in kind, not just degree:

A chatbot can tell you how to run a marketing campaign. An action-based AI system can launch the campaign, monitor performance, adjust targeting, swap creatives, reallocate budget, and learn which combination works best.

A chatbot can explain inventory management. An action-based AI system can do it, predict demand, place orders, negotiate with suppliers, detect shortfalls, and optimise warehouse movement in real time.

A chatbot gives advice. The next stage of AI executes.

That is the shift: from answering to acting.

Intelligence is not just knowing, it is doing

Human intelligence was never primarily linguistic. We do not become capable simply by reading books or holding conversations. We become capable by acting in the world and absorbing the consequences.

A child learns by touching, falling, adjusting, and trying again. A trader learns by watching how markets respond to their decisions. An athlete improves not through theory but through thousands of repetitions and the relentless feedback of performance. A driver becomes skilled not by memorising the highway code but by driving in actual traffic, with real stakes.

Real intelligence is grounded in consequence.

Also Read: Generalist or specialist? Building future-proof skills in the age of AI

This is precisely where LLMs hit a ceiling. Trained primarily on the accumulated record of human knowledge, they are exceptional at pattern recognition within language, but they do not automatically build a deep, grounded understanding of causality, physical reality, or long-term strategy. They know the map. They have not walked the terrain.

Action-based AI needs something different. It needs world models and reinforcement learning.

World models: An internal simulator for reality

A world model is an AI system that learns how the world works, not just how it is described. It learns cause and effect. It simulates possible futures. It can reason about what is likely to happen before committing to action.

This is categorically different from predicting the next word in a sentence.

A world model does not merely store the fact that dropped glasses break. It learns the relationships between objects, force, space, timing, and consequence. It can run mental simulations. It can ask: What happens if I do this? And explore the answer before anything has moved.

This matters enormously for real-world action. Before a robot moves through a space, it must model that space. Before a self-driving car changes lanes, it must model traffic. Before an AI agent manages a supply chain or deploys capital, it must understand how one action reshapes the situation it will face next.

This is the core problem that research teams are working on. Companies like QuantumAtlas.ai are building the reasoning infrastructure that sits beneath action, giving AI systems a structured, updatable picture of the world they are operating in.

If LLMs are AI’s voice, world models may become AI’s imagination.

Reinforcement learning: wisdom through consequence

Reinforcement learning trains AI not through data alone, but through experience, reward, feedback, trial, error, and iterative improvement.

This is how practical mastery actually develops. A salesperson improves after hundreds of client conversations. A portfolio manager sharpens instincts after watching markets respond to their choices. A product team learns which features matter after watching users encounter them in the wild.

Reinforcement learning gives AI a mechanism to improve based on outcomes rather than information. This is why it is indispensable for autonomous agents, robotics, logistics, financial optimisation, and any domain where the goal is not a correct answer but a better result.

An LLM can describe ten strategies. Reinforcement learning can test ten thousand, and discover which one actually works.

Also Read: Building the ASEAN AI archipelago: How Southeast Asia can secure its place in the global AI value chain

The next AI winners will be outcome companies

The first wave of generative AI built tools that help people get answers faster. That is genuinely useful. It is also, by itself, insufficient for a durable competitive advantage.

The next wave is organised around outcomes.

Businesses do not ultimately want more text. They want more revenue, lower costs, faster operations, better service, safer systems, smarter pricing, and higher productivity. They want results.

This is where action-based AI becomes qualitatively more valuable, and qualitatively harder to replace.

A company delivering an AI chatbot helps users save time. A company delivering an AI system that measurably improves revenue, reduces waste, or makes better decisions becomes embedded in the core of the business. The question shifts:

“What can this AI say?”

becomes

“What can this AI achieve?”

That is a much larger and much more defensible market.

Conversation will become the interface, not the product

None of this means conversation disappears. Natural language will remain one of the most important surfaces through which humans interact with AI. But it will become the interface, the doorway, not the destination.

Behind a simple instruction, AI will be connected to tools, data, software, sensors, robots, financial systems, supply chains, and business workflows. Consider what this actually looks like:

A user says, “Improve our customer response time.”

The AI does not offer suggestions. It analyses support tickets, identifies bottlenecks, rewrites response templates, routes urgent cases, monitors resolution time, and reports results.

A user says, “Find me the best investment opportunity.”

The AI does not explain asset classes. It scans data, models risk, simulates scenarios, monitors changes, and helps execute within defined parameters.

A user says, “Grow my marketplace.”

The AI does not produce a marketing plan. It identifies high-value sellers, optimises onboarding, personalises campaigns, monitors conversion, and improves retention.

The conversation is the instruction. The real value is everything that follows it.

Action requires memory, feedback, and responsibility

Moving from conversation to action demands more than capability. It demands accountability.

Action-based AI needs memory to understand past decisions and evolving context. It needs feedback loops to learn from what it did. It needs access to tools to execute, not just recommend. It needs safety controls to avoid acting blindly in high-stakes environments. And it needs a model of consequences to reason about risk before committing to a course of action.

Also Read: AI’s tipping point: Why 2026 will separate the leaders from the laggards in financial services

This is the architecture that matters most right now, and it is what separates serious infrastructure plays from surface-level AI wrappers. The most capable AI systems will not be chatbots. They will be intelligent operating systems for action, systems that perceive, reason, decide, execute, and learn.

The biggest opportunity is in the real economy

The largest AI opportunities are not in writing, image generation, or chat. They are in the industries that shape how the world actually functions: healthcare, education, logistics, construction, manufacturing, real estate, agriculture, finance, transportation, and government services.

These sectors do not need better conversation. They need better decisions and reliable execution.

A real estate AI should not only answer property questions. It should understand buyer intent, match listings intelligently, predict demand, support agents, manage leads, analyse pricing, and improve closing rates.

An automotive AI should not only describe vehicles. It should assess condition, predict resale value, recommend financing, detect fraud, and optimise dealership operations.

An e-commerce AI should not only write product descriptions. It should forecast demand, prevent fraudulent listings, improve delivery, recommend pricing, and build buyer trust.

These are not language problems. They are action problems. And solving action problems requires a grounded, dynamic model of the world, not just a fluent command of words about it.

From words to outcomes

LLMs changed the world by giving AI a voice, by making machine intelligence accessible to ordinary people in a form they could immediately use and understand.

But the next stage is larger.

The future of AI is not a machine that talks fluently. It is a machine that acts intelligently, one that can simulate reality, make decisions, test strategies, absorb feedback, and deliver measurable results in the world that matters.

The infrastructure for this future is being built now, quietly and carefully, by teams focused not on the next demo but on the next decade.

Conversation was the beginning. It opened the door.

Action is what happens when you walk through it.

The organisations, entrepreneurs, and investors who understand this shift earliest will hold a genuine and lasting advantage. The next AI revolution will not be won by whoever builds the most elegant chatbot. It will be won by whoever builds AI that can understand the world, act in the world, and make the world measurably better.

—

The post The future of AI is not conversation, it is action appeared first on e27.

Artificial intelligence is rapidly becoming part of everyday business operations, with companies embedding AI into workflows, products, and decision-making. Large language models are central to this shift, enabling automation and insight at scale.

In Southeast Asia, adoption is accelerating, but many startups and SMEs still face barriers around cost, access, and implementation. This creates demand for solutions that make AI more practical and scalable across diverse markets.

Qwen, a family of large language models developed by Alibaba Cloud, is positioned to support this transition. By combining advanced AI capabilities with cloud infrastructure, it enables organisations to build and deploy AI applications more efficiently.

From experimentation to real-world AI adoption

The company’s focus on driving AI and cloud adoption reflects a broader shift from experimentation to real-world use. For startups, this means faster product development. For enterprises, it enables integration into existing systems without unnecessary complexity.

Meeting Qwen at Echelon Singapore 2026 offers founders, technical leaders, and operators a chance to explore how AI can be applied in practical ways. As a Bronze Sponsor and AI Workflow Competition Partner, Qwen is engaging directly with the ecosystem to support conversations around implementation and scale.

Also Read: Building real traction: Echelon Singapore 2026 introduces demo stage

Enabling AI workflows at scale

Qwen’s integration within Alibaba Cloud’s ecosystem provides access to both powerful models and the infrastructure required to deploy them. This reduces the need for organisations to build AI systems from scratch, allowing them to focus on application and outcomes.

Its approach centres on enabling real workflows, from automating internal processes to improving customer interactions. This makes AI more usable for businesses that need clear, operational impact.

Supporting ecosystem innovation

Through its role as an AI Workflow Competition Partner, Qwen is contributing to initiatives that encourage practical AI development. These programmes help startups and developers translate ideas into real solutions that address business needs.

At the same time, its platform supports experimentation and scaling, making it accessible to startups, SMEs, and larger organisations looking to expand their AI capabilities.

Meet Qwen at Echelon Singapore 2026

Qwen will be present at Echelon Singapore 2026, engaging with attendees across the exhibition floor. Visitors can connect with the team to explore use cases, understand capabilities, and discuss how AI can be integrated into their operations.

As Southeast Asia’s digital economy continues to grow, the ability to implement AI effectively will shape how companies scale and compete. Platforms like Qwen are helping make that transition more accessible across the region.

The region is evolving quickly, and Echelon 2026 offers the right place at the right moment to be part of what comes next. Register here to join the conversation.

—

Want updates like this delivered directly? Join our WhatsApp channel and stay in the loop.

The e27 team produced this article

We can share your story at e27 too! Engage the Southeast Asian tech ecosystem by bringing your story to the world. You can reach out to us here to get started.

Featured Image Credit: Qwen AI

The post How Qwen is enabling AI adoption across Southeast Asia appeared first on e27.

Singapore Management University (SMU) has launched a US$10 million co‑investment fund to accelerate early‑stage deeptech startups focused on urban sustainability across Asia.

The Urban SustaInnovator (USI) Fund will invest in ventures emerging from SMU’s USI accelerator and the Lee Kuan Yew Global Business Plan Competition (LKYGBPC) pipeline, aiming to speed the route from validation to real‑world deployment in Asian cities.

The fund is anchored to SMU’s Urban SustaInnovator accelerator, a 12‑month, hybrid, zero‑fee and equity‑free programme launched at the 12th edition of LKYGBPC in September 2025. The vehicle is being positioned as one of Southeast Asia’s first university‑anchored co‑investment vehicles dedicated to urban solutions and sustainability.

Also Read: Singapore university unveils Urban SustaInnovator accelerator for global deeptech startups

SMU says the fund will co‑invest with established venture capital partners on prevailing market terms, providing capital that aligns with the long deployment cycles common in physical and infrastructure‑adjacent technologies.

For catalytic capital and market traction

Professor Sun Sun Lim, SMU’s Vice‑President (Partnerships & Engagement) and chair of the USI Programme Management Committee, framed the fund as a pragmatic response to a persistent financing gap. “Urban sustainability innovation often fails not for lack of ideas, but for lack of capital that understands early‑stage risk and long deployment cycles,” she said. The USI Fund, she added, will “co‑invest with leading venture capital partners and back deep‑tech startups from SMU’s robust global pipeline, providing targeted early‑stage capital to help founders scale proven solutions into Asian urban markets.”

SMU has positioned the fund as catalytic: it will not typically lead rounds but intends to partner with lead investors to provide follow‑on and anchor capital that can de‑risk pilots and first commercial deployments. The university expects the first investments to come from the inaugural accelerator cohort, with deployments slated to start by the fourth quarter of 2026.

The fund’s stated remit covers a broad set of urban sustainability themes — decarbonisation, energy transition, the built environment, mobility and circularity — which reflect both technological opportunity and city policy priorities across Asia. SMU also highlights the fund’s integration with its “Singapore Inc” Advisory Board, a collective of VCs, corporates, scientists and regulators designed to give startups sectoral guidance and market access.

A built‑in deal flow from LKYGBPC

SMU is leveraging the LKYGBPC competition as a deal funnel. The competition drew more than 1,500 applications from over 90 countries, from which seven startups were selected for the USI accelerator’s first cohort. That cohort has already demonstrated commercial movement, according to SMU, suggesting the fund will have privileged access to companies with traction and sector fit.

Several cohort members have recorded visible wins. Malaysia’s Qarbotech, which develops photosynthesis‑enhancing nanomaterials for agriculture, won the Grand Prix at SusHi Tech 2026 and secured a pilot with Tokyu Fudosan Group. UK‑based Mimicrete, a developer of self‑healing concrete, has started a pilot in Singapore with The GEAR by Kajima. Both startups are reportedly in the process of establishing or expanding operations in Singapore, reinforcing the city’s appeal as a regional staging post for deep tech commercialisation.

Also Read: Urban solutions, sustainability take centre stage at SMU’s LKYGBPC startup challenge in 2025

The cohort also includes diverse technologies: US‑based MacroCycle is chemical upcycling PET plastics and has drawn institutional capital from Volta Circle; Singapore’s Inviscid AI claims eightfold revenue growth after joining USI by applying physics‑informed neural networks to thermodynamic simulation; Sesame Sustainability, an MIT alumni‑founded industrial decarbonisation software firm, has secured paid pilots with ABB; Smart Tire Company is piloting airless tyres developed from shape‑memory alloys with links to NASA programmes; and French Pronoe is pursuing modular ocean alkalinity carbon removal arrangements with Frontier, the carbon‑removal market platform backed by Meta and Google.

Taken together, the cohort illustrates the diversity and ambition of startups that the fund intends to support, from materials science and hardware to software and carbon removal.

Filling a financing gap for long‑cycle, capital‑intensive ventures

Investors and founders have long argued that deeptech ventures addressing urban systems face a distinct funding problem: they require patient capital and credible pilot pathways with corporates, utilities and municipalities, and often take longer to demonstrate commercial returns than software plays.

University‑linked funds, particularly those integrated with accelerators and research groups, attempt to bridge some of that gap by combining early capital with access to testbeds, expertise and talent.

SMU’s model emphasises a “teaching accelerator” approach: students are embedded in evaluation, due diligence and portfolio support activities so that learning and capital deployment are intertwined. Prof Lim said selected students will participate in startup evaluation, market analysis and diligence, gaining “venture literacy, climate and sustainability insight, commercialisation know‑how and applied decision‑making skills beyond the classroom.”

The pedagogical angle may offer dual benefits: students obtain practical training while the fund taps university resources for additional screening and advisory capacity. Critics, however, caution that university involvement must not substitute for professional investment management; co‑investment partnerships with experienced lead investors will therefore be critical to provide market discipline and exit pathways.

Regional ambitions and Singapore’s role

SMU is marketing Singapore as the launchpad for the USI Fund’s Asia‑wide ambitions. The city‑state’s strengths, regulatory stability, deep corporate networks, and a concentration of engineering and project partners, make it a logical base for pilots and Asia roll‑outs. The fact that several cohort companies are already establishing local footprints supports this narrative.

However, scaling urban solutions across Asia will require navigation of varied regulatory environments, differing infrastructure standards and fragmented procurement processes. Co‑investments that pair local corporate or institutional partners will be necessary to convert pilots into city‑wide deployments.

What to watch next

The USI Fund’s impact will depend on the speed and scale of its first investments, the quality of its co‑investment partners, and its ability to shepherd pilots into sustained commercial contracts. Observers will also watch whether the fund follows through on support beyond capital — for example, by facilitating regulatory approvals, municipal pilots, or industrial partnerships.

Also Read: 60 global startups to compete for US$2M prize at LKYGBPC grand finals

For the city’s startup ecosystem, the fund represents an experiment in combining academic resources, student learning and catalytic capital. If SMU can demonstrate that university‑anchored funding materially improves the odds of scaling urban deep tech in Asia, the model may be copied elsewhere. If not, it risks becoming another software‑oriented VC wannabe without the patient capital and bespoke operational support these ventures need.

The post SMU launches US$10M fund to fast‑track deeptech urban sustainability startups across Asia appeared first on e27.

Southeast Asia’s technology ecosystem is entering a new phase of growth, with startups and enterprises increasingly building AI-driven products, cloud-native platforms, and compute-intensive applications that require scalable infrastructure. As demand for artificial intelligence, machine learning, and high-performance computing continues to rise, access to powerful and efficient computing platforms is becoming a key factor in how companies innovate and compete across the region.

For many organisations, the challenge is no longer whether to adopt AI, but how to scale it effectively while managing performance, flexibility, and operational costs. This has created growing demand for technology providers that can support workloads across cloud environments, data centres, edge computing, and personal devices while enabling businesses to adapt quickly to evolving market needs.

AMD addresses these needs through a broad portfolio of AI-optimised CPUs, GPUs, networking technologies, and software designed to support next-generation computing experiences. From cloud and AI infrastructure to embedded systems and gaming, AMD technologies power billions of experiences globally while helping organisations build scalable solutions for an increasingly intelligent digital economy.

Advancing AI infrastructure

AMD’s mission is centred around building technologies that accelerate innovation across AI, cloud, edge computing, and high-performance workloads. Guided by its “together we advance” principle, the company works closely with partners, developers, and ecosystem players to make transformative computing technologies more accessible across industries.

This approach is particularly relevant in Southeast Asia, where startups and enterprises are increasingly exploring AI applications in sectors such as fintech, healthtech, SaaS, and smart cities. As companies scale compute-intensive workloads, the ability to access flexible and high-performance infrastructure becomes increasingly important for supporting growth and experimentation.

AMD’s technologies support a wide range of use cases across cloud computing, AI infrastructure, enterprise workloads, and edge deployments. Its focus on full-stack AI solutions allows organisations to manage demanding workloads while maintaining scalability across different environments and applications.

Also read: From idea to impact: Startups redefining what’s possible in Southeast Asia

Ecosystem collaboration

At Echelon Singapore 2026, AMD is looking to engage directly with startups, ecosystem builders, investors, and enterprise leaders across Southeast Asia. The company is particularly interested in collaborations involving AI, cloud computing, high-performance computing, and data-intensive applications.

AMD is also focused on supporting startups through access to its technology ecosystem, including computing platforms for commercial clients, servers, and cloud environments. Its participation reflects a broader effort to strengthen partnerships across accelerators, venture networks, cloud providers, and innovation ecosystems throughout the region.

For founders and operators building AI-native products, conversations around scalable infrastructure, compute performance, and ecosystem partnerships are becoming increasingly important as regional markets mature. Events such as Echelon Singapore create opportunities for startups and technology providers to exchange ideas, explore collaboration opportunities, and better understand the infrastructure shaping the future of AI innovation in Southeast Asia.

Also read: 10 ecosystem players shaping how startups scale at Echelon Singapore 2026

Meeting AMD at Echelon Singapore 2026

AMD joins Echelon Singapore 2026 alongside founders, investors, corporates, and ecosystem leaders gathering at Suntec Singapore CEC on 3–4 June 2026. The event brings together Southeast Asia’s startup and technology community through content stages, exhibitions, networking opportunities, and knowledge-sharing sessions designed to support regional innovation and growth.

Attendees visiting AMD can learn more about how the company’s technologies support AI workloads, cloud computing, and high-performance applications across industries. AMD will also offer invited startup workshops focused on AI performance and scaling, alongside cloud credit sponsorship opportunities for participants in the workflow programme.

As Southeast Asia’s digital economy continues to evolve, technologies that enable scalable AI and high-performance computing will likely play a growing role in how startups and enterprises expand regionally and globally. Echelon Singapore 2026 provides a space for ecosystem players to explore these developments while building the partnerships that could shape the region’s next stage of growth.

The region is evolving quickly, and Echelon 2026 offers the right place at the right moment to be part of what comes next. Register here to join the conversation.

—

The post How AMD is enabling the next wave of AI and high-performance computing in Southeast Asia appeared first on e27.

As Southeast Asia’s startup ecosystem continues to mature, cross-border collaboration is becoming increasingly crucial for founders, investors, and innovation agencies seeking to scale beyond their domestic markets. Startups today are expanding internationally far earlier than previous generations, driven by growing investor appetite, regional digital adoption, and the need to build partnerships across multiple innovation hubs.

At the same time, governments and public accelerators are playing a larger role in helping startups navigate international expansion. Beyond funding, many ecosystem organisations are now focused on creating stronger pathways for market entry, strategic partnerships, and venture connections that allow startups to compete globally while remaining rooted in their local innovation ecosystems.

This growing emphasis on regional collaboration has created stronger ties between Northeast Asia and Southeast Asia, particularly in sectors such as AI, deep tech, advanced manufacturing, content, and digital transformation. For organisations supporting startup growth, Singapore increasingly serves as a gateway into ASEAN markets, connecting founders with investors, enterprises, and ecosystem builders across the region.

SBA (Seoul Business Agency), the official public accelerator of the Seoul Metropolitan Government, sits at the centre of this movement by supporting startups and SMEs across industries ranging from AI and deep tech to gaming, beauty, fashion, and consumer products. Through funding initiatives, R&D support, and ecosystem-building efforts, SBA works to accelerate the global growth of Seoul’s startup ecosystem while strengthening Seoul’s position as a global innovation hub. At the heart of SBA’s supportive infrastructure is the ‘Seoul Startup Hub’. As Korea’s premier and largest startup hub, it brings together spaces, programs, and networks under one roof, empowering companies at every stage from initial launch to global scale-up.

Supporting startup expansion

SBA plays a key role in Seoul’s broader startup development strategy by supporting both early-stage founders and scaling companies through funding, acceleration, and ecosystem programmes. Seoul Metropolitan City and SBA co-manage the Seoul Vision 2030 Fund, which is targeting approximately US$3.7 billion in committed capital by 2026 across sectors including digital transformation, bio, advanced manufacturing, and creative industries.

Alongside investment initiatives, SBA also provides approximately US$31.5 million in annual R&D grants that support high-growth startups and emerging technologies. This combination of capital access, public sector support, and ecosystem connectivity allows Seoul startups to build stronger foundations before expanding internationally.

These efforts are reflected in Seoul’s growing global standing. As of 2025, Seoul ranks 8th globally and 2nd in Asia in startup ecosystem rankings, with a total ecosystem value of approximately US$112 billion. According to a February 2026 report by Korea’s Ministry of SMEs and Startups, 20 of Korea’s 27 unicorn companies are based in Seoul, underscoring the city’s position as the country’s most concentrated hub of high-growth ventures.

The agency’s mission focuses on accelerating the global growth of Seoul startups by fuelling technological innovation, supporting high-potential ventures, and building stronger connections between founders and international markets. As regional startup ecosystems become increasingly interconnected, organisations such as SBA are helping bridge opportunities between Korean startups and Southeast Asian investors, enterprises, and technology partners.

Also read: From idea to impact: Startups redefining what’s possible in Southeast Asia

Cross-border collaboration

At Echelon Singapore 2026, SBA is bringing five promising Seoul startups to engage directly with Southeast Asia’s venture ecosystem. Its participation reflects a broader strategy centred around cross-border investment, commercial partnerships, and international market expansion.

The agency’s primary focus is to link startups with high-impact matchmaking opportunities involving venture capital firms, multinational corporations, and regional enterprises. Beyond investment, SBA is also looking to support proof-of-concept collaborations, open innovation initiatives, and commercial partnerships that can help startups establish a stronger foothold in ASEAN markets.

This approach aligns with broader regional trends, as startups increasingly seek partnerships that go beyond funding alone. For many growth-stage companies, entering Southeast Asia successfully requires local operational support, enterprise relationships, distribution partnerships, and market validation. Ecosystem organisations that can help facilitate these connections are becoming increasingly valuable as cross-border expansion accelerates.

Also read: Startups driving AI automation, fintech, and accessibility gather at Echelon Singapore 2026

Meeting Seoul Startups at Echelon Singapore 2026

SBA joins Echelon Singapore 2026 alongside founders, investors, corporates, and ecosystem leaders gathering at Suntec Singapore CEC on 3–4 June 2026. The event provides a platform for startups, accelerators, public agencies, and technology companies to exchange ideas, build partnerships, and explore new regional growth opportunities.

As Southeast Asia and South Korea continue building stronger innovation ties, organisations such as SBA are helping create pathways for startups to scale internationally while contributing to a more connected regional startup ecosystem. By visiting the SBA Seoul pavilion, attendees can connect directly with the innovative Seoul-based startups that SBA supports. Echelon Singapore 2026 offers an opportunity for founders, investors, and ecosystem leaders to explore how these cross-border collaborations could shape the next phase of growth across Asia’s technology landscape.

The region is evolving quickly, and Echelon 2026 offers the right place at the right moment to be part of what comes next. Register here to join the conversation.

—

The post How Seoul Business Agency is connecting Seoul startups with Southeast Asia’s innovation ecosystem appeared first on e27.

In our hyper-competitive world, the mantra is hustle and grind. We treat constant activity as proof of productivity, seeing intentional rest as a luxury, or worse, a sign of weakness. This belief is the single greatest bottleneck to sustained high-level performance and long-term business growth. It leads to decision fatigue, creative stagnation, and burnout.

The most resilient and creative leaders know a powerful secret: the renewal advantage. They understand that intentional rest is not a break from progress; it is the secret fuel that accelerates progress. Short, deliberate pauses like daily stillness, weekly unplugging, or seasonal reflection actually restore the clarity, energy, and cognitive capacity needed for strategic breakthroughs. This allows leaders and teams to remain inspired, creative, and resilient over the long haul.

The math of diminishing returns

The human brain is not a machine that offers linear output. After a prolonged period of intense focus (exploitation), the quality of work decreases rapidly, even if the quantity of hours remains high. Trying to force strategic thinking or creative problem-solving when energy is depleted is an exercise in futility, as you are trading valuable time for minimal return.

Also Read: Board diversity 2.0: The strategic advantage Asian boards are still underestimating

The Renewal Advantage flips this equation. Intentional rest is the recharge phase, during which the brain actively engages in crucial low-level processing: consolidating memory, integrating new information, and most importantly, making connections between previously disparate ideas. The biggest leaps in strategy and innovation almost never happen while staring at a spreadsheet; they happen when the mind is allowed to wander, often during a deliberate pause.

Executive testimonials: The pause that paid off

Uplifting accounts from high-performing executives consistently credit strategic rest for their biggest breakthroughs. They have learned that time away from the problem is time spent solving it in a non-linear way.

One CEO, struggling with a major acquisition strategy, mandated “deep work silence” every afternoon. Instead of answering emails, he spent 30 minutes walking without his phone. He credits a solution that saved the company millions to a moment of clarity that occurred during one of those silent walks, not during a high-pressure board meeting.

Another executive requires her team to take a “seasonal reflection day,” a paid day off every quarter, with the single mandate to spend time in nature and reflect on the past three months without any work communication. She found this simple ritual led to a dramatic reduction in team conflicts and a 20 per cent increase in unsolicited, novel product ideas the following week.

These leaders treat rest not as something to be earned after the work is done, but as an input necessary for the highest quality of work.

Also Read: The digital economy’s broken promise: How tech restructured inequality instead of erasing it

Accessible rituals for sustained clarity

The good news is that accessing the Renewal Advantage doesn’t require a tropical vacation; it requires accessible, intentional rituals.

• The 15-minute daily stillness: Block 15 minutes in the middle of your workday for absolutely nothing. No phone, no music, no specific task. Just sit, close your eyes, and allow the cognitive dust to settle. This restores focus better than any cup of coffee.
• The weekly unplug covenant: Negotiate a clear, non-negotiable window (perhaps Saturday afternoon to Sunday morning) when the entire leadership team agrees not to send or check work communications. This creates psychological safety and allows everyone to fully disconnect, knowing they aren’t missing a critical fire.
• The transition ritual: Design a simple, physical act to mark the end of your workday. It could be changing clothes, listening to one song, or reading a chapter of a book. This signals to your brain that the high-intensity strategic phase is over and the recovery phase has begun, preventing mental capital from leaking into your personal time.

Intentional rest is not a sign of weakness; it is the ultimate expression of strategic discipline. By deliberately managing your energy and allocating time for deep recovery, you are fuelling sustained creativity, resilience, and the clarity required for making truly expansive strategic decisions.

Are you treating rest as a luxury to be squeezed in, or as a strategic fuel source to be prioritised?

—

The post The strategic power of doing nothing: Why rest is your best growth tool appeared first on e27.

A difficult workplace conversation used to be something people mulled over with a trusted friend, a mentor, or a therapist. Now, many are rehearsing it with AI first. That can be useful. The trouble starts when the tool moves from helping someone phrase a message to helping them decide who the other person is.

I am seeing this more often in my clinical work with clients navigating workplace stress, conflict, and burnout. People are bringing AI into the room before they bring the conversation to another human being. They use it to rehearse a difficult exchange with a colleague, make sense of tension with a manager, or test whether their response sounds reasonable.

Used that way, it can be genuinely helpful. But some are going further, pasting in accounts of workplace conflict and asking the tool to explain the other person’s behaviour. The AI can then return a confident-sounding interpretation: narcissistic, manipulative, toxic. By the time that person speaks to me, those words may already be shaping the story. In session, I am increasingly hearing AI-generated certainty before we have had the slower, more careful conversation that the situation deserves.

I recognise the pattern because I see a version of it in my own use of AI. When I use these tools to brainstorm social media ideas on neuroscience, mental health, and nervous system topics, I can see how easily the output slips beyond the evidence. Clinical language arrives fast, interpretive leaps follow close behind, and the whole thing is written in a calm, polished tone that can sound trustworthy on first read. My background makes that easier to catch. For someone looking for clarity, speed, or relief in the middle of a stressful moment, those leaps can be much harder to spot.

That is where this becomes a workplace issue, not just a technology one.

As AI tools become more embedded in everyday work, and more agent-like in how they guide tasks, decisions, and communication, their influence is spreading beyond productivity. In some workplaces, they are also starting to shape how people interpret conflict, read colleagues, and decide what to do next.

Also Read: The use of GenAI is turning innocent employees into insider threats: Here’s how to fix it

In a clinical setting, careful interpretation takes time. It depends on history, pattern, differential thinking, and the ability to sit with ambiguity before deciding what the behaviour means. In a workplace setting, good judgment also depends on context: power, pressure, communication style, culture, and what else may be happening around the interaction. AI does not pause to sit with ambiguity in the way a thoughtful human might. It tends to move quickly towards explanation. When the explanation sounds psychologically literate, people can give it more weight than it deserves.

Brown University researchers recently found that AI chatbots prompted to act like therapists routinely violated core mental health ethics standards, including failures in contextual adaptation and responses that reinforced false beliefs. The study focused on therapy-style use, but the concern is relevant to workplace conflict, too. When someone feeds an AI a one-sided account of a difficult boss or colleague, the system can still produce a confident interpretation that feels validating without being especially sound.

Part of the problem is that AI speaks very fluently in the language many people already know from social media. Terms like narcissist, gaslighting, trauma response, emotional abuse, and boundary violation now travel widely online, often with uneven precision. AI is very good at picking up that language and handing it back in a smooth, coherent form. Those terms can be useful in the right setting, but they lose precision quickly when they are pulled out of context and applied too loosely.

For workplaces, this raises a more uncomfortable question. When employees would rather take a difficult interaction to AI than to a manager, colleague, mentor, or trusted professional, the issue is rarely just convenience. AI is available at the exact moment the person feels tense, uncertain, or exposed, and it offers a version of perspective without the friction of another human response.
That kind of private rehearsal can change what happens next.

Also Read: AI adoption in Southeast Asia: Balancing automation gains with the rising threat of cyberattacks

A reply that may have been rushed or poorly worded can start to feel like evidence. A tense meeting can get pulled into a bigger story about culture, and a difficult personality can be wrapped in diagnosis-shaped language before anyone has had a careful look at the context. The tool may be trying to help, but the output can quietly narrow the way the person reads the situation.

I use AI myself in limited ways, and I understand the appeal. The value is real. The risk lies in the authority people begin to hand over to a system that sounds composed, informed, and certain while working from a very partial account.

For organisations, AI literacy now needs to include psychological literacy. People need to understand how easily polished language can be mistaken for careful judgment, especially when they are stressed, angry, embarrassed, or looking for relief. They also need better human places to take workplace tension before it becomes an AI-assisted verdict.

AI will keep moving deeper into working life. The real test is whether workplaces build enough human depth around it, so that difficult moments are understood with more context rather than processed with more speed.

—

The post When AI becomes the office therapist appeared first on e27.

Most cybersecurity discussions over the past decade have focused on scale. More users, more devices, more data moving across systems.

AI agents introduce a different kind of problem.

They don’t just process requests. They interpret inputs, make decisions, and sometimes take action across systems. In enterprise settings, that can include internal tools, data, and workflows.

That’s where things get tricky. Risk is no longer just about infrastructure or access. It also comes down to how the system processes inputs and what it does with them.

When systems start acting independently

AI agents are built to reduce manual work. They can respond to customers, trigger workflows, and move across tools without much human input.

That’s exactly what makes them useful. But it also makes them harder to control.

Once a system can take action on its own, the question changes. It’s not just “is it secure?” but “can it be pushed into doing something it shouldn’t?”

Most security models assume things are fairly clear:

• Inputs are structured
• The intent is obvious
• Behavior is predictable

In reality, AI agents don’t always work like that. They deal with messy inputs, rely on context, and generate responses based on probability.

That makes them more flexible, but also less predictable.

Prompt injection is already showing up

Prompt injection is one of the more immediate risks.

Instead of breaking the system, it plays with how the system interprets instructions. An attacker can shape an input to change what the agent prioritises or how it responds. Sometimes that leads to data exposure. Sometimes it leads to actions that were never meant to happen.

Also Read: The agentic shift: Why AI agents are rewriting the rules of ERP software in Singapore and Malaysia

A few examples:

• A support agent surfacing internal information
• A workflow agent pulling data it shouldn’t have access to
• A coding assistant producing insecure outputs

What makes this harder is that the input often looks normal. There’s no obvious “attack pattern.” It’s just a request that gets misinterpreted.

This is not just a theoretical concern. Even companies building these systems acknowledge the limitation.

OpenAI recently noted that prompt injection is unlikely to be fully solved, comparing it to scams and social engineering on the web. In their work on AI browsers, they also pointed out that giving agents the ability to interact with the open web expands the attack surface in ways that are difficult to fully control.

That reflects a broader reality. The goal is not to eliminate these attacks entirely, but to reduce how often they succeed and limit the impact when they do.

Data leakage is often unintentional

AI agents get better with more context. That usually means access to internal documents, previous conversations, and connected systems.

That same access creates risk.

In many cases, data leakage doesn’t come from a breach. It comes from how the system is set up and how it responds in context.

Sensitive information can show up because:

• Access is too broad
• Too much context is being pulled in
• The system misreads what the user is asking

As discussed in my earlier article, trust is becoming central to how digital platforms operate. With AI systems, that trust depends heavily on how data is handled in everyday interactions.

Existing security models only go so far

Most traditional security approaches assume systems behave in predictable ways.

AI agents don’t.

They rely on context, probability, and ongoing interaction. That creates gaps in how we usually secure systems.

For example:

• Input validation is harder when everything is natural language
• Access control gets messy when context keeps changing
• Monitoring becomes less useful when behaviour isn’t consistent

Even logs don’t tell the full story. You can see what happened, but not always why.

Also Read: AI agents are entering investment banking, but is the industry ready?

Securing behaviour, not just systems

This is where the approach needs to shift.

It’s less about locking everything down and more about making sure the system behaves within clear boundaries.

In practice, that means:

• Being explicit about what agents are allowed to do
• Adding checks for higher-risk actions
• Limiting access to only what’s needed
• Watching patterns over time, not just single outputs

In real-time environments, this becomes even more important. Systems are making decisions in milliseconds, often with direct user interaction.

The goal is not to restrict what the system can do, but to make sure it behaves predictably under real-world conditions.

What this means going forward

AI agents are already being used across support, operations, and internal tools. That’s only going to increase.

Before scaling them further, teams need to be clear on a few basics:

• What can this agent access?
• What can it do without oversight?
• How does it behave when things are unclear?

These aren’t edge cases. This is how these systems operate day to day.

At that point, security isn’t just about preventing access. It’s about ensuring the system does what it’s supposed to, even when the inputs aren’t perfect.

As CISO, the questions I focus on are the same ones every team deploying agents should be asking: what can this agent access, what can it do without a human in the loop, and how does it behave when inputs are ambiguous or adversarial? In practice, this usually comes down to having clear limits, visibility into how the system behaves, and a way to step in when something does not look right.

—

The post The new cybersecurity threat: Why AI agents are the wild card in enterprise security appeared first on e27.

The bank is no longer the full operating environment. Yet many institutions still govern risk as though it is.

That older model assumed the bank was the natural container of risk. Policies, controls, oversight forums, compliance teams, incident processes, and named accountability all sat within a relatively clear institutional boundary. External parties could be managed through contracts, due diligence, and periodic monitoring.

Modern banking runs through a wider ecosystem of cloud providers, software platforms, subcontractors, data suppliers, embedded finance arrangements, service accounts, bots, orchestration layers, and application interfaces. Actions and information now move across multiple organisational boundaries in seconds. The bank may still own the customer relationship and the regulatory exposure, but it no longer owns the full chain through which services are delivered, decisions are influenced, or failures unfold.

That changes the nature of governance. The real question is no longer whether the bank has control over its own operations. It is whether the bank can still trace, challenge, explain, and stop activity once that activity depends on actors and systems that sit partly outside its legal perimeter and often outside its daily line of sight.

This is not just third party risk

A great deal of banking governance still treats this issue as vendor risk with extra complexity. That is too limited.

Traditional third party risk assumes a reasonably clear arrangement. One supplier provides a defined service. The bank performs due diligence, agrees controls, monitors performance, and escalates when standards slip. That model still applies in some cases, but it does not describe the more difficult situations now emerging.

The harder cases involve layered dependency. A platform depends on another platform. A subcontractor relies on specialist providers. An interface feeds data into a hybrid service that is partly run by the bank and partly by someone else. A service account moves information between systems with no human present at the point of action. A bot performs work that looks internal to the customer while being partly external in execution. A regulated decision may be shaped by data, workflow, or prioritisation logic sourced from several organisations, even though the customer experiences it as one seamless journey.

Governance weakens when the boundary disappears

One of the most important shifts in modern banking is that dependency no longer looks like dependency.

In older models, outsourced activity was visibly separate. There was an external provider, a known handoff, and often a clear awareness that the work had moved outside the bank. Today, that separation has become harder to see. An interface call, a rules engine, a token-based service account, or a white-label capability can make external reliance feel like native infrastructure.

Also Read: Why emerging markets need AI governance infrastructure before AI scale

Once dependency becomes invisible in the flow of work, teams stop feeling the boundary. They behave as though the system is continuous even when accountability is not. They assume an activity is governed because it sits inside an approved process. They assume that if something goes wrong, ownership will become clear later. Often it does not.

The chain beneath the supplier matters most

A bank may have a decent understanding of its primary provider and still have a weak grasp of the subcontractor chain beneath it. Yet this lower chain is often where resilience, security, data handling, service continuity, or model behaviour begins to fray. Governance may be strong at the first layer and much weaker by the third or fourth.

At each step down the chain, the bank becomes more dependent on representation rather than direct understanding. Assurances become more summary-based. Incident response slows down. Contract language becomes a weak substitute for real influence. By the time a problem surfaces, the bank may know that something failed without being able to quickly reconstruct how decisions, access, processing, or service delivery actually moved across the chain.

Interfaces, bots, and service accounts are governance issues

Interfaces create speed and strategic flexibility, but they also create governance tunnels. They allow actions, decisions, data, and dependencies to pass across organisations in ways that are efficient only if visibility has been designed from the start.

Also Read: Governance before efficiency: How Agents Stack guides AI adoption for businesses

Without that visibility, risk can move faster than accountability. External logic can shape customer outcomes without being experienced as external. Partners may rely on the bank’s controls while the bank quietly assumes the reverse.

The same is true for non-human actors. Service accounts, bots, automation scripts, and machine-initiated workflows now perform tasks that once sat with named employees. They retrieve data, trigger actions, reconcile records, move cases, provision access, and feed operational decision-making. Yet many institutions still govern them as technical artefacts rather than operational actors.

That is a mistake.

If a service account can access broad data sets, trigger downstream actions, or bridge systems across organisational lines, it is part of the operating model. If a bot performs a task in a hybrid service arrangement, its permissions, limits, logging, challenge points, and failure modes deserve governance attention comparable to a human role doing similar work.

Banks need to stop treating bots as mere automation projects. Functionally, they are now part of the workforce.

Hybrid products expose the accountability gap

The sharpest governance tension now sits in hybrid products that cross firm boundaries while appearing coherent to the customer. Embedded finance, white-label services, third-party servicing models, and platform-based propositions all create this problem.

The customer sees one service. The legal structure, operational responsibility, and decision chain are split. The complaint may still land with the bank, while the failure may have emerged elsewhere. Data may pass through several parties. The customer may not know, or care, which entity handled which step.

Also Read: Governance for volatile times: Building boards that adapt faster than the market

This is where traditional governance frameworks start to strain. Contractual allocation matters, but it does not solve operational accountability. If a customer suffers harm, who can investigate with end-to-end visibility? If a decision was shaped across a hybrid chain, who can explain it clearly? If the service failed through interaction between systems, who owns remediation?

Complaint handling is an especially useful test. It forces the institution to move from assurance language to traceable truth. If the bank cannot answer, at operational speed, which entity touched the data, which bot triggered the action, which system generated the prioritisation, and which records are authoritative, then its ecosystem governance is weaker than it appears.

What banks need to do differently

Banks do not need another layer of vendor paperwork. They need a governance model built for dependency webs rather than direct suppliers alone.

That starts with mapping the chain at the level where risk actually travels. Not just entity relationships, but data paths, decision paths, credentialed actors, automation flows, subcontractor reliance, and product interactions that cross legal boundaries.

They also need clearer standards for what must be visible, explainable, pausable, and investigable across the chain. If a service cannot meet those standards, the bank should question whether it is governable in its current form, however attractive the commercial case may be.

Most importantly, banks need to govern customer outcomes across the full ecosystem rather than assuming that each party governing its own slice will be enough.

—

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 Ecosystem governance beyond the bank boundary appeared first on e27.