AI agents and chat interfaces are no longer limited to answering questions or recommending content. They increasingly act on behalf of users—approving transactions, scheduling actions, filtering information, and making decisions that once required human judgment. This shift is subtle but profound. When systems act for us, cybersecurity is no longer just about protecting data; it becomes about protecting trust.

When automation enters the workflow

In many organisations, AI agents are introduced to improve speed and efficiency. Customer support bots resolve tickets. Financial systems flag or approve transactions. Internal copilots summarise meetings and suggest decisions. At first, these tools feel like assistants. Over time, they become delegates.

The transition often happens quietly. A system that once suggested an action is now executing it. A chatbot that once escalated issues now resolves them autonomously. This is where the security conversation usually lags behind the product decision.

The moment trust becomes a concern

Trust issues tend to surface only after something goes wrong. A transaction is approved that should not have been. An automated message shares sensitive information. A system makes a decision that no one on the team can fully explain.

What makes these incidents different from traditional security failures is diffused responsibility. No single person made the decision. The system did—based on rules, models, and data pipelines built by multiple teams over time.

When users interact with AI through natural language, the system feels human. That perception increases trust, sometimes beyond what the system actually deserves. Users disclose more information. They question decisions less. Attackers understand this dynamic and exploit it.

Also Read: Hunters in the dark: AI agents and the cybersecurity trade-off

Accountability in machine-led decision

AI agents change how accountability works. In human workflows, responsibility is clearer. A person approves a payment. A manager signs off on access. With AI agents, decisions are distributed across models, prompts, APIs, and permissions.

When something goes wrong, teams often ask:

• Was it a data issue?
• A model behaviour?
• A prompt design flaw?
• Or a lack of human oversight?

From a cybersecurity perspective, this ambiguity is a risk. Systems that act autonomously require explicit accountability frameworks, not implicit trust in automation.

New risks introduced by chat interfaces

Conversational interfaces create security risks that traditional systems did not face. Natural language is flexible, ambiguous, and emotionally persuasive. This opens new attack surfaces:

• Prompt manipulation that bypasses safeguards
• Social engineering through AI-generated responses
• Over-permissioned agents that can act across systems
• Users mistaking confident language for correctness

Unlike classic software vulnerabilities, these risks are behavioural. They sit at the intersection of human psychology and system design.

Overconfidence in AI-driven systems

Founders and teams are often overconfident in AI systems because they appear intelligent. A system that explains its reasoning convincingly can mask uncertainty or error. This creates a false sense of security.

Overconfidence shows up when:

• Human review is removed too early
• Audit logs are minimal or absent
• Edge cases are dismissed as rare
• Security is assumed to be “handled by the model”

In reality, AI systems amplify existing risks if governance does not evolve alongside capability.

Also Read: Trust by design: Why cybersecurity is the new economic backbone

Different sectors, different expectations of safety

Expectations of safety vary widely across sectors. In fintech or health, users expect rigorous controls and clear accountability. In media or productivity tools, the tolerance for error is higher until trust is broken.

AI agents blur these boundaries. A general-purpose chatbot used in a low-risk context today may be embedded in a high-risk workflow tomorrow. Security assumptions must travel with the agent, not the use case.

Rethinking responsibility and risk

The key shift is not technical; it is conceptual. Teams must move from asking “Is the system secure?” to “Who is responsible when the system acts?”

This means :

• Designing AI agents with least-privilege access
• Keeping humans in the loop for high-impact decisions
• Logging not just actions, but reasoning paths
• Stress-testing systems for misuse, not just failure
• Training teams to question AI output, not defer to it

Security becomes a shared discipline across product, engineering, and leadership—not a downstream checklist.

One lesson for building teams with AI today

The most important lesson is simple: do not outsource trust to machines.

AI agents can act, decide, and communicate at scale—but accountability remains human. Teams that build secure, trusted AI systems are not those with the most advanced models, but those that design for scepticism, transparency, and responsibility from the start.

As AI agents continue to take action on our behalf, cybersecurity will be defined less by firewalls and more by how well we understand and govern the relationship between humans and machines.

—

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.

The post The new cybersecurity battlefield: Protecting trust in the age of AI agents appeared first on e27.

Many SMEs in Southeast Asia only realise their finance setup is no longer working when growth accelerates. Processes that once felt manageable begin to strain as transaction volumes increase, additional stakeholders enter, and reporting requirements tighten. 

In reality, clearer workflows and stronger process visibility enable problems to be identified much earlier, before delays, errors, and cash-flow blind spots become structural.

The issue is not poor execution. More often, finance systems in SMEs are designed for stability rather than growth.

SMEs power Southeast Asia, but finance is still treated as an afterthought

Small and medium enterprises dominate Southeast Asia’s business landscape. Across the region, SMEs account for roughly 97 per cent of businesses and contribute over 40 per cent of GDP.

Despite their scale, many SMEs still rely on manual, disconnected processes for core finance tasks. Research by the Economic Research Institute for ASEAN and East Asia (ERIA) highlights persistent barriers to digital adoption in developing Asian markets. These include limited business knowledge, gaps in ICT skills, and a lack of localised support.

We see the same pattern in mature markets like Australia. Even with great tech available, an October 2025 OFX report found that 80 per cent of Australian SMEs still rely on manual processes to reconcile expenses. In fact, nearly 38 per cent of business owners report that simple manual data-entry errors are their biggest daily headache. 

It’s a classic case of ‘if it isn’t broken, don’t fix it’ until the manual workload finally becomes too heavy to manage.

Finance issues follow predictable patterns as businesses scale

As SMEs grow, financial complexity increases faster than many teams expect. Invoice volumes rise. Transactions multiply. More people touch the process. Customers and suppliers operate across borders. Regulatory and reporting requirements tighten.

When finance processes are not redesigned for higher volumes, familiar issues begin to surface:

• Invoices are sent late or tracked inconsistently
• Approvals are concentrated with one individual
• Reconciliation is rushed at month-end
• Cash flow visibility becomes limited

These challenges are not surprising. They are the natural outcome of processes that were never redesigned as volumes increased. 

Consulting and software surveys repeatedly point to the same outcome. Weak invoicing and reconciliation processes that depend on spreadsheets or email lead to delayed payments, write-offs, and significant time spent chasing basic financial information.

Also Read: Security, trust, and the future of finance in an AI-driven world

Automation is about reducing friction, not adding tools

Automation is often misunderstood as a large-scale system change or a heavy transformation, when in reality it is primarily about reducing operational risk and manual friction.

For most SMEs, progress starts much smaller.

OECD research on SME digitalisation shows that smaller firms adopt digital tools more slowly than larger organisations, even though the efficiency gains are often proportionally greater. The challenge is rarely technology alone. It is deciding where to start and what to simplify.

In practice, effective automation focuses on removing repetitive friction:

• Standardised invoice workflows
• Automated reminders instead of time-consuming follow-ups
• Approval steps that do not depend on one person
• Fewer instances of entering the same data multiple times

The priority is reliability first. Speed and sophistication follow naturally once the basics are stable.

A practical three-step reset for SME finance

For SMEs looking to improve finance operations, a phased approach is often the most effective.

• Step 1: Make the workflows visible

Document how invoicing, payments, expenses, and compliance actually work today. Simple process mapping often reveals duplication, unclear ownership, and hidden bottlenecks.

• Step 2: Fix the biggest point of friction

Focus on one or two problem areas, such as unpaid invoices, approval delays, or reconciliation backlogs. Small, targeted improvements here often deliver immediate operational and cash flow benefits.

• Step 3: Connect workflows over time

Gradually link invoicing, payments, reconciliation, and reporting so information flows with fewer handoffs. This is where finance shifts from record-keeping to decision support. Research by McKinsey has shown that connected finance workflows can significantly shorten close cycles, in some cases from weeks to days.

Also Read: Why perfect carbon audits could cripple climate finance — and what to fix instead

What consistently works across SMEs

Across SMEs in Singapore and the wider region, several patterns are consistent:

• Small, focused improvements outperform large system overhauls
• Early clean-up reduces operational and compliance risk
• Clear records make audits, fundraising, and reporting easier

When finance operations are stable and predictable, less time is spent fixing errors and more time is available for planning and execution.

The payoff of clear finance processes

Finance rarely becomes a problem overnight. It becomes one gradually, as systems fail to keep pace with growth.

Effective finance operations do not need to be complex. They need to provide dependable visibility. Knowing who owes the business money, what needs to be paid, and where cash stands makes day-to-day operations calmer and month-end faster.

As Southeast Asia’s SMEs continue to expand across borders and operate in increasingly regulated environments, financial maturity will become a competitive advantage rather than a compliance requirement. Clear, connected finance processes provide the operational foundation for sustainable growth and long-term competitiveness.

—

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 Instagram, Facebook, X, and LinkedIn to stay connected.

The post The SME finance reset: 3 steps to fix what’s breaking your growth appeared first on e27.

Artificial intelligence is rapidly becoming the operating layer of the digital economy. Businesses are using AI to automate customer support, improve marketing outreach, and analyse large volumes of data in real time. According to McKinsey, 88 per cent of organisations now use AI in at least one business function, a significant increase from just a few years ago.

Customer engagement is one of the areas changing the fastest. Gartner predicts that conversational AI agents could automate up to 70 per cent of customer interactions by 2027, fundamentally reshaping how companies interact with customers.

Across Asia, this shift is already underway. In Singapore, companies are increasingly using AI across marketing analytics, sales automation, and customer engagement as they look to manage growing volumes of digital interactions. However, as AI becomes embedded in everyday business operations, an important question is emerging. Who actually gets to participate in this AI-powered economy?

The answer will depend not only on access to data or talent, but also on something far less visible. It will depend on the infrastructure that allows businesses to deploy AI systems reliably at scale.

If access to that infrastructure remains limited to large technology companies, the AI revolution could reinforce existing inequalities in the digital economy. But if the tools to deploy AI become easier to access, a much broader range of organisations will be able to build and benefit from AI-powered services.

The infrastructure gap in AI adoption

Much of the global conversation around AI focuses on breakthroughs in large language models. However, turning those models into real-world applications requires far more than simply connecting to an API.

Real-time AI systems often require multiple technologies working together simultaneously. These include speech recognition, natural language processing, text-to-speech synthesis, and networking infrastructure capable of delivering responses instantly.

For many organisations, especially smaller companies and startups, integrating these systems presents a major technical challenge. A Gartner survey found that 85 per cent of customer service leaders plan to explore or pilot conversational AI, yet many organisations still struggle to move from experimentation to full deployment.

One reason is that real-time interactions place strict demands on infrastructure. Even small delays can make AI conversations feel unnatural. Systems must process speech, interpret intent, generate responses, and deliver audio output within milliseconds.

Technology platforms are beginning to address this complexity by combining these components into integrated systems. For example, communications technology provider Agora recently introduced a conversational AI agent solution that integrates speech recognition, large language models, and text-to-speech technologies within a single orchestration layer designed for real-time conversations. 

The platform also relies on a globally distributed real-time network designed to maintain low latency and stable communication across different network conditions. Infrastructure like this aims to remove some of the production challenges that have historically limited voice AI deployment.

Other companies, such as Google Cloud and Amazon Web Services, provide APIs and cloud services that allow developers to embed messaging, voice communication, and AI capabilities into applications without building the entire infrastructure stack themselves. By simplifying these technical requirements, such platforms may help more organisations experiment with and deploy conversational AI.

Also Read: Your biggest competitor might be the AI answer itself

Voice AI and the next interface of digital services

Voice-based AI agents are emerging as one of the most transformative applications of artificial intelligence.

Customer service, sales outreach, and digital support channels are increasingly powered by conversational interfaces that allow users to interact naturally with businesses. Instead of navigating complex menus or typing long messages, users can speak directly with AI systems capable of understanding requests and responding in real time.

This shift is already visible across multiple industries.

Banks are already deploying AI-driven systems to manage financial services and transactions. In Singapore, DBS Bank recently partnered with Visa to pilot Visa Intelligent Commerce, a platform designed to enable secure, agent-initiated payments where AI agents can make purchases or transactions on behalf of consumers with consent and authentication safeguards.

Singapore Airlines recently partnered with Salesforce to introduce AI agents that assist customer service teams by summarising customer interactions and recommending responses in real time, helping staff respond more efficiently during booking inquiries or travel disruptions. E-commerce companies are also exploring conversational and voice-based interfaces to support product discovery, customer support, and post-purchase assistance.

Voice interfaces also offer important accessibility benefits. Speaking is often more intuitive than navigating complex applications, particularly for users who are less comfortable with digital interfaces. However, building voice-based AI systems that feel natural requires extremely reliable infrastructure. Conversations must occur instantly without noticeable delays. Systems must maintain accuracy even in noisy environments or unstable network conditions.

These technical requirements have historically limited large-scale deployment. Platforms that provide real-time communication networks and integrated AI orchestration are attempting to change that by making voice AI easier to deploy across industries.

The future of work in an AI-driven economy

The growth of conversational AI also raises important questions about the future of work.

AI systems are increasingly capable of handling routine customer interactions such as appointment reminders, billing inquiries, and product information requests. Automating these tasks can help organisations manage growing service volumes while improving response times.

However, automation does not necessarily mean replacing human workers.

Also Read: What is zero-click AI visibility? Impact on digital strategy & conversions

Research has shown that AI assistance can significantly improve productivity for employees, particularly when AI helps workers resolve issues more quickly or provides real-time guidance. In customer service environments, AI agents can handle repetitive inquiries while human agents focus on complex issues that require empathy, judgment, or negotiation.

In sales environments, AI tools can assist with lead qualification and outreach while sales teams focus on building relationships and closing deals. Ensuring that workers benefit from this transition through training and new opportunities will be essential to building a more inclusive digital economy.

Equity requires accessible infrastructure

As artificial intelligence becomes embedded in nearly every digital experience, conversations about equity in the digital economy must extend beyond funding and talent.

Infrastructure plays a critical role.

Who has access to the platforms that make AI usable in real-world applications?

Who can deploy AI-powered services quickly and affordably?

And who is excluded when the barriers to adoption remain too high?

The next phase of digital innovation will not be defined only by breakthroughs in AI models. It will also be shaped by the infrastructure that allows businesses of all sizes to turn those models into real products and services.

If these tools remain accessible, the AI era could unlock opportunities across industries and markets. But if access to AI deployment infrastructure becomes concentrated among a few dominant players, the gap between digital leaders and everyone else may continue to widen.

Building equity into the digital economy ultimately means ensuring that the power of AI is not reserved for a select few but is available to the many organisations and innovators shaping the future of technology.

—

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 Instagram, Facebook, X, and LinkedIn to stay connected.

The post Building an inclusive AI economy starts with access to deployment tools appeared first on e27.

At Echelon Philippines 2025, a panel moderated by Twwo Jaruthassanakul of Seedstars explored the evolution of the country’s venture ecosystem.

Speakers Joan Yao of Kickstart Ventures, Joseph de Leon of Manila Angel Investors Network, and Paulo Campos III of Kaya Founders reflected on how the once-fragmented startup landscape has matured into a more cohesive and dynamic ecosystem. Growth has been fueled by diverse talent, including the “sea turtle” phenomenon—Filipinos returning home after studying or building companies abroad.

Despite the regional funding winter, the Philippines has maintained steady investment activity, supported by a vibrant and increasingly organized community of angel investors helping nurture early-stage startups.

The post Echelon Philippines 2025 – Holistic development of the venture ecosystem in the Philippines: From early fragmentation to cohesive growth appeared first on e27.

FinAI has become the essential backbone to enable secure agentic commerce at scale as AI drives change across every part of the economy, said Jiang-Ming Yang, Chief Innovation Officer of Ant International.

Global shifts in commerce and payments

• The speed of consumer AI adoption has outpaced almost any other technology in history, with analysts forecasting AI-facilitated spend to reach nearly US$8 trillion by 2030 — nearly a quarter of all online sales.
• Consumers are embracing new payment methods. Digital wallets and other alternative payment methods (APMs) as well as open banking continue to grow in popularity, especially in emerging markets. Juniper Research expects the number of digital wallet users to grow to over 6 billion by 2030, covering over three quarters of the global population.
• Emerging markets are driving global growth, but merchants there face increasing foreign exchange volatility and high barriers to doing business on international e-commerce platforms.

Also read: Why WorldFirst’s latest move could change how digital platforms scale worldwide

FinAI paving the last mile for next-gen commerce

FinAI will be key in helping merchants navigate global payments systems and adapt to AI-driven commerce, Yang said. Payment firms will become one-stop FinAIaaS partners enabling businesses to engage customers more efficiently, immersively and securely.

According to Yang, Ant International provides five types of critical FinAI capabilities:

• One seamless checkout for cross-channel payments (card, digital wallets, and open banking),
• One agent partner to resolve global payment complexity,
• Customisable solutions for agentic payments and commerce,
• Embedded payments for extra value-added, and
• AI-powered payment security foundation.

Agentic fintech to businesses of all sizes

With AI, technology and operating know-how can be distilled into a single agent, enabling businesses to conduct end-to-end operations from onboarding to optimising payment success rates through one partner. Solutions such as Antom Copilot, which can reduce merchant payment integration time by up to 90% and improve dispute-handling efficiency by 46%, vastly expand access to growth opportunities.

“In the past, only large enterprises had the luxury of hiring large teams to handle the complexities of dealing with global expansion and different payment methods,” said Yang. “Now, AI can change the way we operate by giving businesses access to a single agent partner that is available 24/7.”

Ant International is already working with major players to support agentic commerce growth, collaborating with Google on its Agent Payments Protocol (AP2) and Universal Commerce Protocol (UCP) standards, which guide how agents can operate across the entire shopping journey.

Also read: Eyes on the prize as biometrics reshape everyday payments

Trust as the foundation for growth

Alongside growth potential, AI also brings new challenges to merchants and consumers. Deepfakes, for example, have become a persistent problem. Ant International has developed an advanced anti-deepfake solution, which demonstrate detection rates of over 99%. Yang also highlighted the company’s SHIELD 3-in-1 Transformer model, which is able to identify high-risk transactions with over 95% precision, as key to providing a single trust layer for AI-driven payment security.

“AI-powered threats are no longer just theoretical, they are a reality that we face today. As technologies evolve, one thing does not change – trust will always be the foundation of payments, and will continue to be at the core of our FinAI development journey,” Yang added. He made the remarks in a case study address at The Economist’s Technology for Change conference in March 2026.

—

Enjoyed this read? Don’t miss out on the next insight. Join our WhatsApp channel for real-time drops.

This is a sponsored article produced by the e27 team.

Featured Image Credit: Ant International

The post Ant International: FinAI paving the last mile for agentic commerce appeared first on e27.

Singapore-based Mozark, which helps organisations check how their digital services actually perform for real users in the real world, has raised US$40 million in Series B funding.

The round was led by IFC (World Bank Group) and RMB Capitalworks, with Kalaari Capital also participating.

Mozark sells what is essentially a reality check for digital products. Its platform runs scripted user journeys on real devices, across real networks and locations, then turns the resulting telemetry into a diagnosis of where performance breaks — from the app layer down to network infrastructure.

Also Read: Transformation tenet: The digital customer experience is key to “stickiness”

The company says it now works with more than 50 enterprise and government customers across 20 countries and has executed more than 25 million tests on several thousand live devices.

Why this matters for Southeast Asia

For Southeast Asia, the significance of the round goes beyond another sizable cheque in Singapore. It reflects a broader shift in the type of digital infrastructure attracting investment.

The region’s digital economy has grown rapidly, but reliability remains highly variable outside premium urban corridors. When apps slow down or fail, the consequences extend beyond user frustration. They can mean missed payments, failed logins, dropped telehealth calls, or unreliable access to government services.

Mozark’s proposition is straightforward: measure digital performance as experienced by real users, not as reported by dashboards inside a cloud region. In markets where regulators, telcos, and critical service providers need evidence of service quality across diverse geographies, that distinction matters.

It also helps explain why IFC’s involvement is notable. For development-focused investors, tools that measure digital reliability are increasingly viewed as part of the economic plumbing of emerging markets, rather than simply another DevOps layer.

Where the new capital will go

Mozark plans to use the fresh capital to accelerate expansion beyond Southeast Asia and deepen its technical capabilities.

The company says the funding will support:

• Expansion into priority markets, including the United States and the Global South
• Strategic acquisitions
• Deeper testing and measurement across what it calls the “AI-native stack”, spanning applications, networks, and AI infrastructure
• Development of agent-to-agent communication testing, designed for systems where AI agents — not just humans — exchange requests and execute tasks

According to founders Kartik Raja and Fabien Renaudineau, the need for real-world testing is growing as digital services become more complex.

“AI is accelerating digital services everywhere, but experience quality remains disparate and unreliable,” they said in a joint statement.

Mozark’s Chief Product Officer Chandra Ramamoorthy points to the underlying constraint: traditional testing approaches still depend heavily on controlled environments. “Testing remains constrained by physical infrastructure limitations,” he said, positioning Mozark’s real-device approach as a way to validate performance at scale under real-world conditions, rather than relying solely on lab simulations.

A market shifting from monitoring to proof

Digital experience monitoring has long relied on dashboards and synthetic checks running from data centres. But in the Asia Pacific, the core challenge is increasing variability.

Users frequently switch between Wi-Fi and mobile networks, rely on mid-range Android devices, and access services that traverse a complex chain of CDNs, telco routing, cloud regions, and third-party APIs.

This complexity is pushing enterprises toward tools that can answer more practical questions:

• How does this app behave on a specific handset model in a second-tier city?
• Is latency caused by the app itself, the CDN, the ISP route, or local congestion?
• Can regulators or enterprises independently verify performance claims?

In other words, the market is shifting from monitoring systems to providing user experience.

A growing but still fragmented market

Public analyst breakdowns typically group “digital experience monitoring” within broader application performance monitoring (APM) and observability markets.

Also Read: How Southeast Asian brands are reimagining the future of digital experiences

By those measures, Southeast Asia remains a relatively small slice of Asia Pacific spending, though growth is accelerating as banks, telcos, superapps, and governments digitise more workflows.

In practical terms, that places the regional opportunity in the hundreds of millions of US dollars annually, with further upside as AI-driven services increase the cost of outages or degraded experiences.

Mozark is positioning itself in the gap between traditional application monitoring, which often assumes stable infrastructure, and network measurement tools, which rarely capture full application journeys.

This positioning may prove particularly relevant in markets where sovereignty-ready deployments and independent verification are becoming increasingly important.

A competitive global arena

Mozark is entering a competitive field populated by well-funded incumbents and specialised measurement platforms.

Key players include:

• Dynatrace, Datadog, New Relic, and AppDynamics (broad observability and APM platforms)
• Catchpoint and ThousandEyes (Cisco) (internet and network experience monitoring)
• Akamai and Cloudflare (performance infrastructure with measurement capabilities)
• Network and mobile performance specialists such as Ookla and Opensignal

Mozark’s differentiation lies in combining real-device, real-network testing across multiple geographies with an emphasis on independent measurement and deployments designed to meet regulatory and data sovereignty requirements.

A rare Southeast Asian contender

Within Southeast Asia, many vendors provide QA testing or performance monitoring. But few homegrown platforms focus on large-scale, real-world device telemetry across multiple countries, serving both enterprises and regulators.

As a result, Mozark often finds itself competing with global platforms or with in-house monitoring solutions that struggle as systems become more AI-driven and interconnected.

The new funding gives Mozark the runway to prove that its model can scale globally.

The company’s broader bet is that digital experience will soon need to be measured as rigorously as uptime, not simply marketed.

The post Mozark raises US$40M to test how apps really behave in the wild appeared first on e27.

The cryptocurrency market advanced 2.15 per cent to reach a total capitalisation of US$2.44T on March 13, 2026. This gain stands out because it occurred while traditional risk assets faced severe pressure. Equities and bonds sold off sharply as Brent crude oil surged above US$100 per barrel for the first time since 2022. Escalating Middle East tensions and a critical blockage in the Strait of Hormuz triggered the move.

The crypto market’s weak correlation with the S&P 500 at -14 per cent and with Gold at -34 per cent signals a crypto-specific catalyst rather than broad risk-on sentiment. This divergence suggests digital assets are beginning to trade on their own fundamental narratives. Such independence represents a maturation I have long argued is essential for the asset class to evolve beyond a speculative adjunct to traditional finance.

The primary engine behind this rally is BlackRock’s launch of its iShares Staked Ethereum Trust, ticker ETHB, which debuted on Nasdaq on March 12. The product generated US$15.5M in first-day volume, a solid start for a novel instrument. This ETF allows investors to gain exposure to Ethereum’s price while simultaneously earning staking rewards. The design treats ETH as a productive, yield-bearing asset. This marks a profound shift.

For years, institutional adoption focused on Bitcoin as digital gold, a store of value. BlackRock’s move validates Ethereum’s utility as a foundational technology capable of generating cash-flow-like returns. By locking up ETH supply through staking, the product mechanically reduces sell-side pressure. This creates a favourable supply-demand dynamic. The critical metric to watch now is weekly ETF flow data. Sustained inflows would confirm that institutions are not just testing the water but are committing capital to this new yield-bearing crypto thesis.

Supporting this institutional momentum is a wave of regulatory optimism. Social media channels buzzed with reports that President Trump had confirmed a zero per cent tax on crypto transactions. Additional chatter highlighted the US Senate advancing measures to block a Central Bank Digital Currency until 2030. While these developments require official verification, the market is clearly pricing in a more accommodating policy environment. This narrative has fuelled a healthy rotation of capital into altcoins. The Layer 1 sector advanced 1.58 per cent.

Artificial intelligence tokens like Render surged over 11 per cent. Bitcoin dominance held steady at 58.78 per cent. This indicates that new money is flowing into the broader ecosystem rather than just fleeing to the largest asset. Such breadth is a positive sign for market health. It suggests investors are gaining conviction in specific technological narratives like decentralised compute and scalable infrastructure.

Also Read: Why crypto surged while stocks fell: The regulatory breakthrough changing everything

From a technical perspective, the market cap is now testing a pivotal level at US$2.44T. Immediate resistance sits at the recent swing high of US$2.46T. A clean break above this level could open a path toward the US$2.52T extension. Caution is warranted because the seven-day Relative Strength Index reads 74.39. This indicates overbought conditions in the short term.

The rally may need to consolidate before its next leg higher. The key support level to monitor is US$2.33T. A break below this floor would signal a loss of momentum and could trigger a deeper pullback. The next major catalyst will be the upcoming US ETF flow reports. Positive data could provide the fuel needed to overcome resistance. Disappointing flows might exacerbate a technical correction.

This crypto-specific rally gains additional significance when viewed against the backdrop of traditional market turmoil. On March 12, US indices posted broad declines. The Dow Jones Industrial Average fell 739.42 points, or 1.56 per cent, to close at 46,677.85. The S&P 500 dropped 103.22 points, or 1.52 per cent, to 6,672.58. This marked its lowest close since November. The Nasdaq Composite slipped 404.15 points, or 1.78 per cent, to 22,311.98 as technology stocks grappled with rising yields. The VIX volatility index settled at 24.23, reflecting elevated fear. The trigger for this selloff was the energy crisis. Brent crude surged over nine per cent to settle at US$100.20 per barrel.

The International Energy Agency warned of the largest oil supply disruption in history. This shock has forced traders to scrap expectations for Federal Reserve rate cuts in 2026. Soaring energy costs threaten to reignite inflation. Consequently, US Treasury yields are climbing. The 2-year yield jumped 11 basis points. The 10-year yield hit 4.27 per cent. Stress is also emerging in the US$1.8T private credit market. Funds like Morgan Stanley and Cliffwater LLC have capped withdrawals following a surge in redemption requests.

In this environment, crypto’s decoupling is not just a market curiosity. It represents a potential shift in how digital assets function within a diversified portfolio. My view has consistently been that crypto’s long-term value proposition hinges on its ability to offer uncorrelated returns driven by its own adoption cycles and technological progress. The current action supports that thesis.

The rally is fuelled by a structural product innovation from the world’s largest asset manager and a favourable regulatory narrative. It is not driven by a surge in liquidity from traditional markets. This is a more sustainable foundation for growth. Sustainability remains the key question. Can the crypto market maintain its upward trajectory if ETF inflows decelerate this week or if the macro backdrop worsens? The overbought RSI suggests a pause is likely. The underlying drivers remain intact.

Also Read: Crypto market surges to US$2.38T as Middle East tensions ease: What comes next

The path forward hinges on a few clear factors. First, institutional demand for the new staked Ethereum ETF must prove durable. Second, the regulatory narrative needs to translate into concrete policy actions to maintain confidence. Third, the market must successfully digest its overbought condition without breaking below the US$2.33T support. A failure on any of these fronts could lead to crypto re-correlating with traditional risk assets. Those assets are currently under severe strain from inflation fears and geopolitical instability. For now, the momentum is bullish, and the drivers are specific to the crypto ecosystem. This is a sign of maturation.

The market is beginning to trade on its own merits. This development aligns with the vision of a decentralised financial system operating in parallel with, and sometimes independently of, the legacy system. The coming days, with their focus on ETF flows and key technical levels, will provide crucial evidence on whether this independence can be sustained amid a global macro storm. Investors should watch the US$2.46T resistance and US$2.33T support as decisive boundaries.

A break above US$2.46T could accelerate gains toward US$2.52T. A drop below US$2.33T would signal a loss of momentum and invite a deeper correction. The US$15.5M debut volume for ETHB offers an initial benchmark, but sustained weekly flows will determine if institutional appetite remains strong.

With Bitcoin dominance at 58.78 per cent, the market retains room for altcoin expansion if the regulatory tailwinds persist. The 7-day RSI at 74.39 warns of short-term exhaustion, so patience may reward those waiting for a healthier entry point. In a world where Brent crude trades above US$100 per barrel and the 10-year yield touches 4.27 per cent, crypto’s ability to post gains on its own terms signals a new phase of market evolution. This phase demands careful monitoring of ETF data, technical levels, and policy developments. The US$2.44T market cap represents both opportunity and risk. Navigating this landscape requires discipline, clarity, and a focus on the structural forces shaping the next chapter of digital finance.

—

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 Instagram, Facebook, X, and LinkedIn to stay connected.

The post Bitcoin and Ethereum rally while S&P 500 plummets: Is crypto finally decoupling from traditional markets? appeared first on e27.

For many startups and SMEs in Southeast Asia, growth beyond domestic markets is often seen as the ultimate milestone. Yet while strong demand signals and product-market fit can spark the ambition to go global, international expansion requires far more than a good product. It demands operational discipline, regulatory awareness, and the ability to scale internal systems across borders.

Karen Ng, Regional Head of Expansion and Market Lead for Singapore, Hong Kong, ASEAN, and India at Deel, says founders frequently underestimate the operational complexity of expanding internationally.

“Going overseas is the ultimate stress test of whether what works at home still works across borders,” she explains in an email interview with e27.

For SMEs hoping to succeed globally, the difference between smooth growth and costly setbacks often comes down to preparation.

Many founders treat expansion as a market problem: if there is demand, expansion should follow. However, operational readiness is often the more critical factor.

In a single country, founders may handle payroll, HR, and compliance manually. Once companies enter multiple markets, that approach quickly becomes unsustainable.

Also Read: Ant International: FinAI paving the last mile for agentic commerce

Administrative responsibilities multiply, especially when navigating unfamiliar labour laws, tax rules, and employment regulations. Without systems in place, founders can quickly become overwhelmed by fragmented processes and compliance risks.

Operational readiness for international expansion, Ng says, should include clear visibility into total workforce costs, repeatable hiring processes, and a reliable system for managing contracts, onboarding, and payroll across markets.

Rather than approaching regulation, hiring, infrastructure, and demand as separate checklists, SMEs should design an operating model that can be replicated in each new market.

In other words, expansion works best when it is built on scalable systems rather than improvised solutions.

Common pitfalls when hiring international talent

Hiring internationally is often one of the first steps in global expansion. However, many Southeast Asian founders mistakenly assume hiring abroad works the same way as hiring locally.

In reality, labour laws, tax registrations, and employment protections differ significantly across jurisdictions. Without proper oversight, companies risk misclassifying workers, failing to register as required, or unknowingly violating employment regulations.

These issues often remain hidden until companies undergo due diligence during fundraising or face regulatory scrutiny.

Another common mistake is assuming recruitment ends once the offer letter is signed. Winning international talent is only the first step. Without strong onboarding systems, clear communication structures, and consistent payroll practices, remote hires may feel disconnected from the organisation.

Also Read: Building an inclusive AI economy starts with access to deployment tools

Companies that successfully retain global talent focus on designing a complete employee lifecycle—from hiring and onboarding to compensation and engagement—ensuring international employees feel integrated rather than peripheral.

Choosing the right workforce structure

For SMEs entering new markets, deciding how to structure their workforce is another key decision.

Founders typically choose between three options: engaging contractors, establishing a local legal entity, or using an Employer of Record (EOR).

Contractors may be suitable for short-term, project-based work with minimal dependency. However, they should not be used as a workaround to avoid employment obligations. If contractors function like employees, companies risk misclassification penalties later.

Setting up a legal entity offers greater control and is often the right approach when a company has strong confidence in a market and plans to hire long-term. However, this route requires managing registrations, payroll compliance, and ongoing regulatory responsibilities in each country.

For many early-stage market tests, Ng says an Employer of Record provides a practical middle ground. An EOR allows companies to hire employees compliantly without establishing a local entity, enabling them to validate market potential before making deeper commitments.

Over time, companies can transition from EOR arrangements to their own entities as expansion matures.

Cybersecurity risks grow with global teams

As SMEs expand internationally, their cybersecurity exposure also increases.

Each additional market, vendor, or HR tool adds new layers of risk. What once may have been a single payroll or HR system can quickly become a patchwork of local providers, spreadsheets, and disconnected software platforms.

Also Read: Building an inclusive AI economy starts with access to deployment tools

The result is fragmented data storage across jurisdictions—often containing sensitive employee and payroll information.

A single security breach or compromised device could therefore trigger regulatory issues in multiple countries simultaneously.

To reduce risk, many companies are shifting towards unified platforms that integrate HR, payroll, and IT access management into a single system. This approach makes it easier to enforce consistent security standards and monitor vulnerabilities across markets.

High-growth startups often prioritise speed, but rapid expansion without governance can create hidden risks.

Compliance, AI governance, and cybersecurity cannot be treated as secondary concerns that are addressed later. Instead, they must be embedded into the company’s operating model from the start.

Ng notes that companies expanding successfully tend to build unified systems for hiring, payroll, and access management rather than rebuilding policies in each new market.

When HR, payroll, and IT processes operate within a single framework, companies gain greater visibility over workforce decisions, data access, and compliance requirements.

Preparing for a more regulated future

Looking ahead, SMEs should expect increasing scrutiny around artificial intelligence, cybersecurity standards, and workforce governance.

Regulators are beginning to examine how algorithms influence hiring and performance decisions, while cybersecurity requirements are tightening as remote work becomes more common.

Rather than focusing on individual tools, regulators are increasingly assessing whether companies have a consistent governance framework across all markets.

For Southeast Asian SMEs pursuing international expansion, the lesson is clear: global growth is no longer just about entering new markets. It is about building an operating model that can adapt to evolving regulations, technologies, and workforce structures.

Companies that design for that flexibility today will find it far easier to scale tomorrow.

The post Why operational readiness is key to successful international expansion for SMEs appeared first on e27.

GoTo’s latest financials tell a familiar story in Southeast Asia’s digital economy: impressive growth paired with lingering questions about profitability.

On paper, the numbers look strong. Gross Transaction Value surged, revenue climbed steadily, and adjusted EBITDA soared past guidance. The standout performer was clearly fintech, where consumer lending and AI-driven underwriting are powering rapid expansion. In many ways, this reflects a broader regional trend—superapps increasingly leaning on financial services as their most scalable and profitable engine.

Yet the headline growth masks a more complicated reality. GoTo still posted a net loss for the year, even as it highlights record adjusted EBITDA. Because adjusted EBITDA excludes several key costs, it offers only a partial picture of financial health. For investors, the real question is whether operational improvements can translate into consistent bottom-line profit.

The quiet sidelining of Tokopedia in headline metrics also suggests a strategic shift. E-commerce remains one of the most competitive and capital-intensive segments in Southeast Asia, and GoTo appears to be focusing attention on higher-margin areas such as fintech and mobility.

Ultimately, GoTo’s results illustrate the delicate balancing act facing regional tech giants: sustaining rapid growth while finally proving that scale can deliver durable profitability.

Regional

GoTo’s fintech arm shines but losses persist in 2025 financials: GoTo posted 57% GTV growth and record adjusted EBITDA in Q4 2025, but a US$95M full-year net loss and heavy reliance on non-IFAS metrics raise questions about true sustainable profitability.

JAFCO Asia rebrands as JIF Capital after Bee Alternatives acquisition: Singapore-headquartered JAFCO Asia has rebranded as JIF Capital following its acquisition by Bee Alternatives, retaining its leadership team and Asia-focused investment strategy across AI, cybersecurity, and enterprise tech.

ByteDance eyes US$2.5B Nvidia Blackwell AI systems in Malaysia: ByteDance is deploying around 36,000 Nvidia B200 chips in Malaysia via Aolani Cloud for AI R&D outside China, in a deal potentially worth over US$2.5 billion amid surging global AI compute demand.

Bukalapak revenue jumps 46% on gaming segment growth: Indonesian e-commerce firm Bukalapak posted US$384M in 2025 revenue driven by its gaming segment, while sharply narrowing adjusted EBITDA losses and holding over US$1 billion in cash reserves.

Custa raises US$4.3M to scale customisation platform across SEA: Malaysia-Japan customisation startup Custa secured pre-Series A funding to expand across Singapore and Southeast Asia, deploying AI across its supply chain after delivering over 500,000 products.

Indonesia restricts social media access for 70 million children: Indonesia’s Komdigi ministry will enforce new regulations from March 28, restricting social media access for children under 16 across platforms including TikTok, Instagram, and YouTube.

Interviews & Features

Skills intelligence is the future of SME hiring: LinkedIn: LinkedIn’s Elsie Ng says AI-assisted hiring tools like Hiring Pro help Southeast Asian SMEs surface stronger candidates faster, with skills intelligence emerging as the next competitive edge.

Zicy’s Alvin Koay: AI visibility is ASEAN’s next competitive frontier: Zicy co-founder Alvin Koay explains why ASEAN startups and MSMEs must urgently optimise for AI answer engines or risk losing customer discovery, trust, and revenue to AI-invisible competitors.

Twilio: AI must bridge digital intelligence and real-world outcomes: Twilio’s APAC VP says winning in Asia Pacific requires AI that bridges digital intelligence with real-world outcomes, powered by scalable infrastructure, voice capabilities, and localised customer engagement strategies.

JFrog: AI workflow tools are an enterprise security blindspot:  JFrog’s security research VP explains how critical sandbox escape vulnerabilities in n8n expose the growing risk of AI workflow platforms being trusted with privileged enterprise access without adequate security controls.

Echelon PH: How startups convert market signals into winning products: Founders from Expedock, Qrospay, and inDrive share lessons on hyperlocal market entry, aligning teams around clear value propositions, and distinguishing genuine demand from misleading market signals.

International

Bangladesh emerges as South Asia’s next smart investment destination: With 170M people, 130M internet users, and a maturing startup ecosystem, Bangladesh’s demographic momentum and valuation affordability are attracting serious investor attention across fintech and digital infrastructure.

Korea’s Dongguk University unveils AI Buddhist robot monk:  South Korea’s Dongguk University has introduced Ven. Hyean, a semi-humanoid AI robot trained on Buddhist scriptures to provide spiritual guidance, counselling, and temple assistance using on-device AI.

Chinese banks ramp up tech lending as Beijing pushes AI agenda: Chinese state and joint-stock banks are shifting lending priorities toward AI, semiconductors, and advanced manufacturing startups, with one bank targeting 30% growth in high-tech loans for 2026.

OpenAI’s DOD deal faces congressional scrutiny over AI warfare: Sam Altman faced tough questions from US senators over OpenAI’s Pentagon contract, with lawmakers demanding guardrails around AI use in kill chains, autonomous weapons, and mass surveillance.

HSBC and StanChart set to receive Hong Kong’s first stablecoin licences: HSBC and a Standard Chartered joint venture are poised to receive Hong Kong’s first stablecoin licences by March 24, marking a major step in the city’s regulatory framework for fiat-pegged digital assets.

Metaplanet launches VC and management units to back bitcoin ecosystem: Tokyo-listed Metaplanet has launched two subsidiaries deploying US$25.2M into Japan’s bitcoin financial infrastructure, targeting lending, payments, custody, stablecoins, and early-stage founders.

Australia’s teen social media ban undermined by VPNs and workarounds: Three months after Australia banned under-16s from social media, teens are bypassing restrictions via VPNs and age verification workarounds, as at least 14 countries consider similar laws.

Cybersecurity

AI-powered cyberattacks are now a structural risk for startups: CrowdStrike’s 2026 report warns attack volumes rose 89% as AI industrialises cybercrime, leaving cloud-native startups dangerously exposed to credential theft, supply chain attacks, and sub-30-minute breaches.

Ad fraud is quietly draining APAC’s travel marketing budgets: Bots account for up to 80% of invalid traffic for travel advertisers in APAC, with Singapore and Vietnam among the most fraud-prone markets, costing brands millions in wasted ad spend.

Malaysia has 72 cyber startups but barely any dedicated funding: Despite ranking among SEA’s top three cybersecurity startup hubs, Malaysia’s ecosystem remains underfunded, with investors favouring generalist tech plays over dedicated cyber infrastructure and platform companies.

Cybersecurity failures destroy trust — and that’s a board-level problem: As AI-powered deepfakes and ransomware attacks escalate globally, cybersecurity must shift from an IT concern to a boardroom priority to protect organisational trust, finances, and reputation.

AI workflow tools are becoming a dangerous enterprise security blindspot: JFrog researchers uncovered critical sandbox escape flaws in n8n, exposing how AI workflow automation platforms trusted with privileged enterprise access are becoming high-value targets for attackers.

Semiconductor

Asia’s quantum hardware race is a geopolitical and economic bet: Quantum hardware attracts 70% of global quantum funding, and Asian nations are building sovereign qubit capabilities, local supply chains, and state-backed R&D programmes to secure long-term strategic advantage.

India plans US$10.8B fund to turbocharge domestic chipmaking: India is preparing a US$10.8B semiconductor fund covering chip design, manufacturing equipment, and supply chain development as it races to build world-class chipmaking capacity by 2032.

South Korea’s semiconductor exports surge 175.9% to record high: South Korea’s semiconductor exports hit a record US$7.6 billion in the first ten days of March, driving overall exports up 55.6% year-on-year and accounting for 35% of total shipments.

AI

Shared intelligence, not shared data: The future of AI collaboration: A new compute-to-data model combining federated learning, differential privacy, and executable governance enables organisations to collaborate on AI without exposing sensitive data across regulatory boundaries.

Tourism is APAC’s most underrated AI opportunity: APAC’s complex, high-volume tourism sector presents a compelling but overlooked AI opportunity, where practical applications reducing operational friction offer far more value than flashy tech demos.

Google Maps gets Gemini-powered Ask Maps and immersive navigation: Google Maps is rolling out a Gemini-powered conversational search feature and upgraded 3D immersive navigation with smart lane guidance, natural voice directions, and personalised route recommendations.

Thought Leadership

Bear market funding: What investors really want now: Investors now prioritise profitability, capital efficiency, and strong teams over hype, with 24–36-month runways and realistic valuations becoming non-negotiable in bear market pitches.

SaaS exit reality: Proceeds matter more than valuation: Founders chasing headline M&A valuations often overlook debt adjustments, earn-outs, share swaps, and post-closing liabilities that dramatically reduce actual cash proceeds at exit.

Asian boards must govern for volatility, not stability: Traditional governance models are too slow for Asia’s volatile landscape. Boards must adopt agile oversight, real-time risk dashboards, and continuous scenario planning to stay strategically relevant.

SEA travel tech must prioritise human connection over transactions: Southeast Asia’s travel startups are removing friction to restore human connection, but risk creating over-curated digital bubbles that strip away the serendipity that makes travel meaningful.

Multi-currency wallets and travel cards redefine cross-border payments: As millennials drive demand for digital-first financial solutions, multi-currency wallets and travel cards are emerging as the most efficient tools for seamless, budget-friendly cross-border travel payments.

RedDoorz: How SEA startups must swap burn rates for profitability: Its CEO argues SEA startups must abandon hypergrowth for sustainable scale, sharing how deliberate market exits, automation, and profitability focus delivered the company’s first positive earnings in 2024.

What sports tech entrepreneurs can learn from travel innovation: A sports investor argues travel’s innovations in biometrics, dynamic pricing, AR/VR, and personalised experiences offer a ready-made playbook for transforming the fan experience in sports venues.

Crypto rose 0.64% while stocks fell: The regulatory shift driving it: As geopolitical tensions rattled equities and gold, crypto’s 0.64% gain to US$2.39T revealed its growing decoupling from traditional markets, driven by a landmark White House pro-crypto policy pivot.

SEA’s next startup wave must be built on identity, not just speed: As SEA’s startup ecosystem matures, founders must anchor growth in philosophical clarity, treasury discipline, and collaborative ecosystems rather than hypergrowth narratives and competitive optics.

Four travel tech startups reimagining SEA tourism post-pandemic: Korean travel tech startups Tripbtoz, Stayfolio, ONDA, and Infoseed are leveraging XR, AI, and precision mapping to redefine post-pandemic travel experiences across Southeast Asia and beyond.

Gen Z’s solo travel boom is fuelling Asia’s travel app market: With 83% of APAC smartphone users having travel apps installed, Gen Z’s preference for solo, experience-driven travel is creating massive opportunities for Asian travel tech startups like Traveloka and Klook.

The post Ecosystem Roundup: GoTo fintech leads growth amid losses | LinkedIn bets on skills intelligence for SMEs | Malaysia’s cyber startups face funding drought appeared first on e27.

Singtel’s venture arm is putting real money behind a thesis many telcos have been flirting with for years: AI is moving from “nice demo” to core infrastructure, and whoever controls deployment pathways (networks, data centres, cloud, and systems integration) gets to shape where the value lands.

On Thursday, Singtel Innov8 unveiled a US$250 million AI Growth Fund aimed at backing growth-stage AI startups globally while pushing applied AI deeper into its telco parent’s own businesses: from customer operations and network management to cyber security and IT automation. The new vehicle sits alongside Innov8’s existing US$250 million evergreen fund, taking total managed capital to US$500 million.

Why this matters for Singtel, and for Singapore, and SEA

For Singtel, the fund is less about chasing the next flashy model and more about industrialising AI inside a telco group. That means turning AI into a repeatable capability across:

Also Read: Building an inclusive AI economy starts with access to deployment tools

• Singtel Singapore and Optus (as “live” testbeds for customer engagement, network operations and automation)
• Digital InfraCo, including Nxera data centres and Singtel’s RE:AI cloud platform (as the compute and infrastructure layer)
• NCS (as the systems integrator to roll out AI for enterprises and governments)

In other words: Singtel is trying to stitch together a full stack — connectivity + compute + deployment — and use venture investing to keep its pipeline of tools and talent topped up.

For Singapore, the significance is symbolic and practical. Symbolically, it reinforces the city-state’s ambition to be a credible applied AI hub, not just a place that hosts regional HQs. Practically, more capital directed at growth-stage AI firms can increase the odds that startups choose Singapore as their regional launchpad, especially if they can pilot products with a major operator and its enterprise customers.

For Southeast Asia, the signal is that large incumbents are no longer waiting for AI winners to emerge elsewhere. Telcos sit on distribution, data exhaust (with heavy governance constraints), and mission-critical operations, and they are under pressure to defend margins. A dedicated AI fund is a way to buy options on the future and bring in technology that can cut costs, reduce churn, and build new enterprise revenue lines.

Innov8 CEO Edgar Hardless framed it in deployment terms: the fund will invest “with clear deployment pathways” so Singtel can “test, integrate and scale AI innovations across our networks, platforms and digital infrastructure.”

What’s the objective of the fund?

Strip away the corporate phrasing and the objective looks like this:

• Find AI companies that can be deployed inside Singtel’s operating companies (not just held on a cap table).
• Speed up adoption by turning Singtel’s assets — networks, cloud, data centres and enterprise integration — into a scaling engine for portfolio startups.
• Target domains that map to telco pain points and enterprise demand, including customer engagement, network operations, cyber security, IT automation, enterprise AI platforms, and vertical AI applications.
• Singtel Innov8 also pointed to alignment with Singtel’s broader infrastructure push, including an “AI Grid” concept spanning 5G-Advanced, edge and cloud, orchestrated via Paragon.

Is this the first AI fund launched by a telco in the world?

Telcos globally have long run venture units and innovation funds that invest heavily in AI-related startups, even if they are not always branded as a standalone “AI fund”. Operators and telco-backed venture arms in the US, Europe, and Asia have been making AI bets for years across security, network automation, enterprise software and data platforms.

Also Read: Envisioning the future: The critical challenges and opportunities of AI investment

What does stand out here is the explicitly AI-labelled, growth-stage focus and the size (US$250 million) coming from a Southeast Asian operator at a time when many corporates are still experimenting with smaller AI budgets and proofs of concept.

How many AI companies has Singtel invested in so far?

Singtel Innov8 said it has made over 120 investments globally since it was established in 2010. However, it is not clear how many of those are specifically AI companies, nor how many new AI investments it expects this fund to make.

What it did disclose: since Singtel’s strategic reset in 2021, Innov8’s investments have generated an internal rate of return of 26 per cent, a figure that will inevitably raise expectations for this new AI vehicle, given the sector’s hype cycle and valuation swings.

AI adoption is accelerating in SEA; Singapore is a key test case

AI adoption across Southeast Asia has shifted from experimentation to competitive necessity, and Singapore is one of the region’s fastest-moving markets because the ingredients are unusually concentrated:

• Government push and policy clarity: Singapore’s national AI programmes, public-sector digitisation and clearer governance frameworks reduce friction for enterprise adoption.
• Dense enterprise and regulated-industry base: Banks, telcos, logistics firms, and government-linked organisations have budgets, and urgent use cases in fraud, service automation, compliance and operations.
• Infrastructure readiness: data centres, cloud penetration and high-quality connectivity make it easier to run AI workloads reliably.
• Labour and productivity pressure: businesses are turning to AI to offset hiring constraints and rising costs.
• Maturing buyer behaviour: companies increasingly want measurable outcomes (time saved, tickets reduced, fraud caught), not “AI theatre”.

Regionally, the drivers look similar but play out differently by market. Indonesia’s scale pushes AI adoption in commerce, customer service and risk; Vietnam and Malaysia are building momentum in manufacturing and shared services; Thailand’s large consumer economy makes personalisation and service automation compelling.

Across the board, cheaper access to models and tooling has lowered the barrier to shipping AI features, while competition has raised the penalty for standing still.

Also Read: AI is eating the world and startups are riding the infrastructure wave

Singtel’s bet is that a telco group with infrastructure, enterprise channels and government relationships can be more than a buyer of AI; it can be a deployment platform. Whether that turns into durable advantage will depend on execution: picking startups that survive the hype cycle, integrating them without suffocating them, and proving that “AI Grid” ambitions translate into hard results in production.

The post Singtel launches US$250M AI fund to turn its telco empire into an AI deployment platform appeared first on e27.