Singapore organisations rank among the most rigorous in the Asia-Pacific region when it comes to cybersecurity governance, yet a significant gap persists between policy intent and operational enforcement, according to new findings from JFrog’s 2026 Software Supply Chain Security State of the Union report.

The report, which surveyed 1,508 IT professionals across eight countries — including 174 respondents based in Singapore — paints a nuanced picture of a market that has invested heavily in governance frameworks but lacks the tooling to make them self-enforcing.

The findings arrive against a backdrop of escalating global threats. Malicious packages uploaded to the npm registry surged 451 per cent year-on-year to 171,592, while 495 weaponised AI models were detected on public registries. A further 11.7 million new packages entered software supply chains over the same period.

On several headline measures, Singapore performed well. The country led all eight surveyed nations on network proxy enforcement, with 67 per cent of organisations applying controls at that layer. Additionally, 71 per cent of Singapore respondents said they carefully review AI-suggested code fixes before implementation — the highest rate of AI scrutiny recorded in the dataset.

However, those strengths are offset by a series of structural vulnerabilities.

Also Read: Why Singapore’s deep tech founders need more than good science — and how National GRIP is filling that gap

Audit readiness emerged as a particular concern. While 95 per cent of Singapore organisations claim to track application ownership, 54 per cent said they would need a week or more to produce compliance documentation for a single application on demand, suggesting that data exists in principle but is not structured for rapid retrieval.

Open-source software approval processes also lag the region. Some 59 per cent of developers in Singapore wait a week or more for new package approvals to be granted, the slowest rate recorded across the APAC markets surveyed.

The report also identified a notable blind spot around so-called “shadow AI tools”. Eighteen per cent of Singapore organisations have formal policies prohibiting the use of unauthorised AI tools, yet have no technical mechanism to detect when those policies are violated — the highest “policy-only” rate in APAC.

Secrets detection, a control designed to identify exposed credentials and API keys embedded in code, remains significantly underdeployed. Only 25 per cent of Singapore organisations have adopted the capability, a figure broadly in line with the global average of 28 per cent.

Human review cannot match development speed

The report highlights the operational strain created by relying on manual processes to govern AI-accelerated development workflows. Sixty per cent of Singapore DevSecOps stakeholders identified security governance and policy enforcement as their primary time burden, while 41 per cent cited the review and hardening of AI-generated code as a significant drain on resources.

Also Read: The talent reset: Why AI is changing what makes people valuable

Sunny Rao, senior vice president of Asia-Pacific at JFrog, said the findings reflected a common transition point for mature markets.

“Singapore has done a lot of hard work in building governance frameworks that most markets are still debating,” Rao said. “Policies that rely on manual review and human checkpoints cannot keep up with AI-driven development. The organisations that will lead from here are the ones that embed enforcement directly into the pipeline — so that every artefact, every model, and every dependency is curated, scanned, and validated before it ever reaches a developer’s machine.”

JFrog’s report points to automated, platform-level enforcement as the recommended path forward — including pre-vetted package curation, automated secrets scanning, and contextual vulnerability analysis to prioritise remediation efforts based on actual deployment environments.

The full report is drawn from JFrog’s global survey of 1,508 IT professionals conducted across eight countries in 2026.

The post Singapore leads on security governance but struggles to enforce it, report finds appeared first on e27.

Bitcoin faces intense downward pressure, tumbling over 6.09 per cent in a 24-hour window to US$66,867.60. This sharp correction means the premier digital asset is notably underperforming the broader cryptocurrency market, which itself fell by 5.39 per cent over the same period.

The velocity of this descent suggests that a complex interplay of excessive leverage, cooling institutional appetite, and structural liquidations has fundamentally transformed what could have been a standard market correction into a disorderly retreat. For observers tracking these capital flows, this pronounced vulnerability highlights how deeply the cryptocurrency ecosystem remains bound to sudden changes in market sentiment and leveraged positioning.

At the very core of this steep price drop sits a massive derivatives-driven liquidation cascade that completely transformed the market dynamics over a 24-hour period. Leveraged long positions worth nearly US$789 million were completely wiped out, forcing automated and non-discretionary market selling that triggered a painful feedback loop. This volume of liquidations represents an astonishing 172 per cent surge over the prior day, proving that retail and institutional traders alike were positioned far too aggressively on the long side.

As prices breached psychological support levels at US$70,000 and US$68,000, exchange liquidation engines automatically dumped collateral into an increasingly illiquid spot market, thereby amplifying the decline’s velocity. Market stabilisation now depends entirely on whether funding rates and open interest can stabilise, signalling that this aggressive excess leverage has been thoroughly cleared from the system.

Also Read: Bitcoin down 3.32% as US$283M in liquidations wipe out leveraged traders: Saylor’s power?

Compounding this structural selling pressure is a visible erosion of institutional confidence, a pillar that many believed would permanently anchor prices throughout the year. For the 11th consecutive day, spot Bitcoin exchange-traded funds registered persistent capital withdrawals, with total aggregate outflows reaching US$3.45 billion.

This prolonged streak of capital flight indicates a broader risk-off rotation as professional allocators quietly shift their capital out of digital assets and reallocate it directly into outperforming traditional equities, with artificial intelligence stocks attracting the vast majority of this liquidity.

Furthermore, sentiment suffered a sharp psychological shock following reports that MicroStrategy executed its first Bitcoin sale since 2022. Even though the transaction was minor, it shattered the firmly held market narrative that the corporate treasury would exclusively accumulate and never sell its holdings, introducing an element of doubt that further spooked already nervous participants.

Bitcoin has officially pushed deep into oversold territory, with its 14-day relative strength index collapsing down to 29.09. The digital asset is currently testing a critical floor between its recent swing low of US$66,127 and the 78.6 per cent Fibonacci retracement level located at US$67,300. If buyers fail to defend this crucial US$66,127 mark, the structural bearishness will likely intensify and open up a direct path toward US$64,000.

Conversely, if exchange-traded fund outflows finally begin to slow or turn neutral today, a successful defence of this support zone could easily spark a quick relief rally back toward the 50 per cent Fibonacci level at US$68,868.

Also Read: Why US$73,000 is the most important Bitcoin level right now

The current assessment points to a market dominated by strong bearish momentum, where the combination of aggressive liquidations and a cooling institutional bid has firmly handed control over to the sellers. While deeply oversold conditions frequently precede a sharp technical bounce, any near-term recovery will likely remain incredibly weak and highly vulnerable until the asset reclaims and stabilises above its key overhead resistance zones.

Risk managers must keep a vigilant eye on today’s exchange-traded fund data to see if institutional selling pressure is showing signs of exhaustion, while simultaneously watching whether the spot price can successfully defend its current support lines.

Evaluating the probabilities for how this market structure will evolve over the coming days reveals three distinct tactical scenarios for risk allocators to monitor. The primary scenario carries a 65 per cent probability and envisions Bitcoin staging a modest technical bounce from its current oversold conditions to retest US$68,868, before ultimately succumbing to the overarching bearish trend and resuming its decline toward US$64,000. There is a secondary 25 per cent probability that the selling pressure has already peaked, which would allow the asset to firmly establish a long-term bottom right here and begin a sustained, grinding recovery that targets a full reclaim of US$70,000.

Finally, a minor 10 per cent probability exists for an immediate, catastrophic continuation of the liquidation event, a worst-case scenario that would bypass any intermediate consolidation and plunge the asset straight through US$64,000 down to deeper macro support levels.

Let’s see.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post A 65% probability explains the next likely move for Bitcoin as leverage clears appeared first on e27.

Cool Japan Fund’s latest disbursement into Jakarta-based vending operator JumpStart has propelled the company from Series B to Series C, reinforcing a bet on automated retail as a conduit for Japanese F&B brands into Indonesia, and by extension, Southeast Asia.

The undisclosed follow-on investment from CJF comes as JumpStart reports rapid expansion of its machine network and remarkable year-on-year revenue gains. The company said it operated more than 6,500 vending machines in Indonesia at the end of 2025 and logged financial growth of up to 200 per cent compared with the previous year.

Also Read: What Japan and Southeast Asia teach us about co-creating innovation

A distribution platform, not just machines, JumpStart positions itself as more than a vending-machine operator. Its strategy is to build a scalable distribution platform that connects international, primarily Japanese, brands with Indonesian consumers via modern retail infrastructure.

CEO Brian Imawan framed the funding as validation of that vision. “The continued support from the Cool Japan Fund is a strong validation of our vision and execution. We are not only building a vending machine network, but also creating a scalable distribution platform to connect global brands, especially from Japan, with the growing Indonesian consumer market,” Imawan said in a statement.

That language is telling: the company is pitching automated retail as both a logistics channel and a marketing front. Machines are placed in high-footfall settings (office towers, transit hubs, universities, and retail precincts) and equipped with AI and cashless payments to enable dynamic merchandising and targeted promotions.

Tech and cashless payments as growth levers

JumpStart’s machines deploy artificial intelligence to optimise inventory decisions and personalise the consumer experience, the company says. Combined with cashless payment systems, the platform can gather transaction data and customer insights that are scarce in many parts of Indonesia’s fragmented retail landscape.

This data-driven approach is part of a wider trend in the region where retailers, digital platforms and payment networks are converging. For international brands seeking market entry, access to local consumer behaviour and real-time performance metrics reduces uncertainty. In markets like Indonesia, where e-wallets and digital payments have leapfrogged in many urban areas, cashless vending reduces friction and expands monetisation opportunities beyond single transactions, for example, through loyalty programmes and push promotions.

Why Cool Japan is doubling down

For CJF, a government-supported vehicle created to promote Japanese culture and industries abroad, the rationale is two-fold. First, Indonesia is a critical consumer market in Southeast Asia: its population exceeds 275 million, and urban middle-class consumption is growing. Second, vending machines offering Japanese snacks, beverages and convenience items are a soft-power instrument, a familiar way to introduce Japanese brands, flavours and lifestyles to a foreign audience.

Also Read: East meets Southeast: How Japan can empower a new wave of SEA startup innovation

CJF’s investment therefore serves both commercial and cultural aims: it accelerates market access for Japanese SMEs and strengthens bilateral cultural ties through everyday consumption.

Operational priorities and challenges

JumpStart intends to use the new capital to expand its machine footprint, broaden its product variety, especially through partnerships with Japanese suppliers, and reinforce its operational infrastructure and supply chains. That mirrors common scaling priorities for automated retail players: site acquisition, restocking logistics, cold-chain management for chilled items, and machine maintenance.

However, scaling a vending network in Indonesia presents specific challenges. The archipelagic geography complicates logistics; ensuring timely restocking and refrigeration across islands increases costs. Site selection remains a labour-intensive process that requires local relationships and negotiation with property owners. Machines also face vandalism and theft risks in some locations, pushing operators to invest in surveillance, sturdier hardware and insurance.

Moreover, the sustainability of growth will depend on the economics of each machine: average revenue per machine must justify hardware, installation, and ongoing operational expenses. The 200 per cent year-on-year growth headline is striking, but sustaining high growth rates as the base expands is statistically harder.

A Southeast Asian playbook

While the current focus is on Indonesia, the broader implication is regional. If JumpStart can prove unit economics at scale and develop reliable supply chains, the model could be replicated across Southeast Asia, especially in neighbouring markets with similar consumer profiles and rising digital payments, such as the Philippines, Vietnam and Thailand.

For Japanese brands, a regional vending network offers a controlled, low-barrier entry point before committing to full retail distribution deals. For JumpStart, the value proposition is to act as both distributor and experiential marketer, a company that can place a product in front of millions of consumers while delivering data on how it performs.

What this means for incumbents and investors

The deal signals investor appetite for niche, tech-enabled retail infrastructure that combines hardware, software and cross-border partnerships. It also raises the bar for incumbents: traditional distributors, convenience store chains, and local FMCG players may feel pressure to match the immediacy and data feedback loop that vending platforms promise.

For investors, the opportunity hinges on the operator’s ability to scale efficiently and to translate user data into higher-margin services, such as targeted advertising, subscription models, or white-label distribution for international brands.

An important route for Japanese brands

CJF’s move to back JumpStart further underlines the convergence of trade, culture and technology in Southeast Asia’s retail sector. The investment is a concrete example of how soft-power funds can catalyse commercial ventures that export culture through consumer goods, and it positions automated vending as a potentially important route for Japanese brands to reach fast-growing consumers in Indonesia and the wider region.

Also Read: “SEA + Japan is a long game”: MUIP’s Gerrard Lai on cross-border startup collaboration

Whether JumpStart can sustain its growth while solving the logistical and economic puzzles of large-scale vending across an archipelago remains the key question. But for now, the deal marks a notable vote of confidence in automated retail’s role in Southeast Asia’s evolving consumer landscape.

The post Cool Japan backs JumpStart’s Series C as vending machines reshape Indonesian retail appeared first on e27.

SG Enviro, a Singapore-headquartered industrial water and wastewater engineering firm, has closed a Series A round led by Emerald Ventures with co-investment from SEEDS, the investment arm of SG Growth Capital.

The funding will accelerate the company’s expansion across Southeast Asia and deepen its presence in Singapore, Malaysia, and Indonesia, with plans to enter Thailand and target high-growth sectors such as data centres, semiconductors and pharmaceuticals.

Also Read: SG Enviro bags US$5.92M to tackle Southeast Asia’s wastewater challenge

The raise comes as regional manufacturers and heavy industries face tightening regulations, rising utility costs and growing corporate pressure to reduce water and carbon footprints. For Southeast Asia, where water stress, ageing treatment plants and resource-heavy industries such as palm oil and oil and gas are concentrated, the timing is significant.

From retrofits to new utility stacks

SG Enviro, founded in 2018, positions itself as a full lifecycle partner: engineering, procurement and construction (EPC), plus operations and maintenance (O&M). Its stated strength is retrofitting and upgrading legacy wastewater infrastructure, an approach that can extend asset life and reduce both environmental impact and operating costs.

Retrofitting older plants is a pragmatic play in the region. Many industrial wastewater systems across the region were designed when environmental expectations were lower; upgrading these systems can deliver immediate compliance and operational gains without the capital outlay of greenfield projects. For industries such as palm oil, food and beverage, and oil and gas –where effluent composition and volumes vary widely — tailored technical solutions are often required, a fact SG Enviro highlights as a competitive advantage.

Targeting higher-margin, high-demand sectors

Beyond traditional industries, SG Enviro is explicitly moving to address the utilities needs of data centres, semiconductor fabs and pharmaceutical manufacturers. These sectors are both water‑intensive and increasingly bound by strict environmental and reliability standards.

Also Read: Innovating for impact: A better solution for household water treatment

For Singapore and Malaysia, countries actively courting hyperscale data centre investment, specialist wastewater and water-reuse solutions can be a differentiator for sites seeking long-term operating licences and investor favour.

Dr Helge Daebel, head of its water practice at Emerald, said. “SG Enviro’s advanced wastewater treatment expertise plays an important role in supporting Singapore’s growing industrial sector while driving its regional expansion. He described SEEDS as a partner with “strong local support capabilities and extensive Southeast Asian reach” that can add value to SG Enviro’s growth.

Investor backing and regional strategy

Emerald Technology Ventures is a global venture firm with a long track record in climate and sustainability-related investments. The firm manages and advises assets totalling over €1 billion (~US$1.1 billion) across offices in Zurich, Toronto and Singapore. SEEDS, operating under SG Growth Capital, co-invests to catalyse private capital into Singapore-based technology startups with global ambitions.

Christine Giam, partner at SG Growth Capital, reinforced the Singapore-to-region narrative: “Leveraging Singapore’s strengths as a hub for engineering innovation and industrial expertise, SG Enviro combines water treatment technologies with deep execution capabilities to address complex industrial wastewater challenges.” Her comment underscores a common playbook: use Singapore as a launchpad for regional scaling, blending engineering credentials with local market partnerships.

Commercial realities and competitive landscape

While the market opportunity in Southeast Asia is clear, competition is intense. Multinational EPC firms, established regional engineering houses and emerging specialist technology providers all compete for industrial water projects. SG Enviro’s focus on retrofits and on-the-ground O&M aims to capture recurring revenue streams that pure equipment providers may miss. Securing long-term service contracts with industrial clients would bolster margins and create stickier customer relationships.

Another practical lever for growth will be strategic partnerships. The company said it plans to use part of the new capital to strengthen its talent pipeline and form alliances—moves that could ease market entry in countries where local content, licences and sector connections matter.

Regulation and corporate sustainability as tailwinds

Southeast Asian governments are increasingly tightening industrial effluent and discharge standards and nudging industries towards circular resource use. At the same time, multinational buyers and lenders are pushing suppliers to demonstrate environmental performance across scope 1–3 emissions and water stewardship policies. These dual pressures make wastewater treatment not just an environmental obligation but a business continuity and finance-enabling service.

For example, palm oil mills and food processors that can demonstrate water reuse and energy recovery through anaerobic digestion can improve margins and access export markets subject to stricter sustainability requirements. Data centres seeking reliability assurances in water-scarce jurisdictions need resilient, fit-for-purpose utilities—an addressable niche for specialist firms.

What to watch next

Investors and clients will be watching several indicators: whether SG Enviro can win larger EPC contracts outside Singapore, the pace of O&M contract wins that deliver recurring revenue, and its ability to form local partnerships in Malaysia, Indonesia and Thailand. The company’s move into highly regulated sectors such as semiconductors and pharmaceuticals will demand both technical depth and compliance credentials.

Also Read: Hydroleap revolutionises wastewater treatment, leading industries into a sustainable future

With Southeast Asia’s industrial base transitioning under regulatory and commercial pressures, there is clear demand for firms that can combine technology with execution. SG Enviro’s Series A gives it financial headroom to test whether that combination can scale across the region.

Possible near-term milestones to look for include announcements of anchor client contracts in the new sectors, local joint ventures or partnerships in Indonesia and Thailand, and the hiring of senior country leads to accelerate sales and project delivery.

The post SG Enviro closes Series A, targets water- and energy-hungry data centres and fabs in SEA appeared first on e27.

For years, the conversation around AI has revolved around one question: Will AI replace jobs?

It is understandable. Change creates uncertainty, and few technologies have moved as quickly or visibly as artificial intelligence.

But this may be the wrong question. The more interesting one is: how does AI change the structure of work itself?

From what I am seeing as a founder building and working alongside AI systems, AI is not simply replacing teams. It is redesigning how work is organised, delegated, and executed. The future workforce may not be defined by company size or headcount alone. Instead, it may increasingly be shaped by how effectively humans collaborate with AI. And that collaboration may look less like software and more like teammates.

Leverage existed before AI — AI made it conversational

Before AI became mainstream, businesses were already using systems and technology to create leverage. We were highly workflow-driven long before AI entered the picture. Processes, systems, and automation were already central to how we operated.

But traditional automation still required technical setup and rigid logic. AI changed something fundamental. It made conversational leverage. Instead of navigating complex systems or relying entirely on human execution, work could increasingly be delegated through natural language.

That shift matters. Because AI did not invent operational leverage. It made leverage more accessible, adaptive, and personalised.

This became especially clear while building Seraphina, my AI twin and assistant. Seraphina was not designed as a generic chatbot. It was built around years of workflows, communication patterns, content, and operational context. What emerged was not simply an AI tool. It felt more like an operational companion. And that experience changed how I thought about work.

Sometimes we are not thinking anymore, we are processing

One of the biggest misconceptions around productivity is that humans spend most of their time thinking. In reality, many knowledge workers spend enormous amounts of time processing emails, coordinating, making drafts, making repetitive decisions, and performing endless operational tasks.

At some point, the workload becomes so heavy that creativity and strategic thinking begin to disappear. You are no longer innovating. You are simply reacting.

Also Read: The AI productivity gurus are bluffing too

This is where AI becomes transformative. Not because it eliminates human contribution, but because it redistributes cognitive effort. When repetitive execution shifts toward AI, humans regain space for strategy, judgment, leadership, relationship-building, creativity, systems thinking, and innovation.

That distinction matters. Because productivity is not only about doing more. It is about creating more room to think better. In many organisations today, the real bottleneck is not intelligence. It is processing overload.

The rise of one person + one AI

This is why I increasingly believe we are moving toward a one-person + one AI workforce model.

Just as many employees today have laptops, email accounts, and productivity software, future workers may have something else: their own personal AI teammate. Not merely a chatbot, but an AI layer that understands personal workflows, communication style, task history, operational preferences, contextual memory, and recurring responsibilities.

We describe this as one person + one AI. The purpose is not to remove humans from work. It is to elevate them.

If execution can increasingly be supported by AI, humans gain more time for higher-value contributions. This naturally shifts organisations away from measuring visible busyness and toward something more meaningful: outcomes. Businesses do not survive because they follow a perfect process. They survive because they create meaningful outcomes. AI simply makes outcome-oriented work more achievable.

I recently shared this philosophy with an intern. My expectation was not that she would spend every hour grinding through tasks manually. The AI could generate much of the execution. Her role was to audit, review, and ensure quality before work went live. The goal was not time spent. The goal was responsible output.

One AI is not enough — the future may be AI crews

Personal AI is only one layer. The next evolution is what I call an AI crew.

Many people imagine AI as one super assistant doing everything. I do not think that is how this develops. Because no single human department handles every function, and no single AI should either.

A founder may increasingly work alongside an AI writer, an AI researcher, an AI marketer, an AI operator, an AI developer, and an AI auditor. Not one AI – a crew of specialised systems. This mirrors how organisations already function. Marketing does not replace finance. Operations does not replace legal. Likewise, specialised AI systems trained around particular workflows often perform more effectively than a single general-purpose assistant.

I have noticed this myself. Seraphina is a strong assistant, but she cannot fully audit herself. Strangely, this resembles human behaviour. Humans have blind spots. So do AIs. Which means future AI systems may increasingly work together, checking, validating, and supporting one another. The future may not be one super-intelligence. It may be coordinated intelligence.

Also Read: Building with AI has never been easier, just do not build the next Chegg

AI will replace some roles, and we should be honest about that

Some roles will likely be reduced or reshaped, especially those centred heavily around repetitive execution, coordination, or predictable processing. But technological evolution has always changed execution.

When Photoshop first appeared, design required specialised expertise and significant training. Then the tools became easier. Canva expanded access. The work changed. Human creativity did not disappear. Execution evolved. AI may follow a similar path.

And while some tasks become automated, human value may shift upward – toward judgment, taste, leadership, creativity, emotional intelligence, strategic thinking, and relationship building. People still buy from people. That has not changed. What changes is how much repetitive labour is required before humans can create that value.

Prompting may be a leadership skill

One surprising insight: people who struggle to delegate to AI often struggle to delegate generally. That is not criticism. It is an observation.

Prompting is frequently treated as a technical AI skill. I see it differently. Prompting is often a leadership and delegation skill. If instructions are vague, unclear, or inconsistent, AI performs poorly. Humans do too.

The reason experienced operators may gain disproportionate leverage from AI is not simply because they use better software. It is because they already understand systems, delegation, workflows, briefing, and operational clarity. AI accelerates these strengths. It also exposes weaknesses faster.

This is why AI alone is not the unfair advantage. Experience plus AI is. A new founder has access to the same tools. But founders who spent years building systems, leading teams, and learning through execution may compress timelines dramatically. What once took months can increasingly happen in weeks. Not because the work disappears. But because execution becomes amplified.

Also Read: Faster tech, slower brains: The biological blind spot of the AI race

Human judgment still matters

Despite everything AI can do, it should not replace human judgment, because not every decision is an efficiency problem. Some decisions involve wellbeing, leadership, values, emotional impact, and sustainability.

Recently, my team declined a client opportunity. Not because the work was impossible – AI could likely have helped process it. But we recognised something more important: the engagement would create mental strain and disrupt team balance. AI might have calculated feasibility. Humans considered consequences. That distinction matters.

AI can optimise processes. Humans remain responsible for deciding what deserves their energy. That is leadership. And leadership still requires judgment.

The workforce may be redesigned, not reduced

AI may give us something many people have quietly lost: time. Time to think. Time to build intentionally. Time to create. Time to reconnect with work that feels meaningful.

The future workforce may not be defined by company size or by who has the largest headcount. It may instead be defined by something far more interesting: how intelligently humans and AI collaborate together. Not humans versus AI, but humans, personal AI, and AI crews working alongside one another.

And perhaps that is the real shift worth paying attention to.

The post Why one person + AI is becoming a serious workforce model appeared first on e27.