Day: August 12, 2026

The digital asset ecosystem is facing a reality check as total crypto capitalisation drops 1.24 per cent to US$2.18T over the past 24 hours. This decline reflects a profound shift in investor sentiment rather than a mere technical correction. Market participants now view digital coins through a strictly macroeconomic lens. They act as highly sensitive barometers of global economic health and liquidity conditions.

My perspective centres on the undeniable fact that virtual currencies now march in lockstep with traditional risk instruments. When broader economic indicators flash warning signs, capital quickly flees speculative ventures. This risk-off reaction highlights the sector’s maturation. Institutional capital dictates the flow and demands alignment with traditional financial metrics.

A surprisingly weak employment report served as the primary catalyst for this broad risk aversion. The United States economy lost 23,000 jobs in July. This figure directly contradicted analyst expectations for job growth. This negative surprise immediately altered the calculus for participants evaluating interest rate trajectories. Weak economic metrics typically prompt expectations of monetary easing. This dynamic explains the strong 66 per cent correlation between the digital asset sector and the S&P 500.

Investors treat both asset classes identically during periods of economic uncertainty. Furthermore, the ecosystem exhibits a negative correlation of 71 per cent with Gold. Traders actively sell risk instruments to buy traditional safe havens when macroeconomic publications disappoint. This clear divergence from precious metals proves that digital tokens currently function as high-beta technology stocks rather than digital gold.

Bitcoin experienced an even steeper decline. The leading cryptocurrency fell 1.97 per cent to US$63,902.70. The premier digital coin underperformed the slightly softer broader environment due to its intense sensitivity to small-cap equities. Bitcoin currently maintains a massive 94 per cent correlation with the Russell 2000 index. This staggering statistical link reveals that allocators view the leading digital coin as a proxy for speculative small-cap stocks.

When economic anxiety rises, participants rapidly dump these high-volatility positions. The shared macro-driven move indicates that fundamental crypto narratives take a back seat to broader economic jitters. Geopolitical tensions in the Middle East further compound this anxiety. These global conflicts force liquidity providers to widen spreads and reduce exposure ahead of critical inflation metrics.

Also Read: Crypto’s new threat is not a hack, but a knock at the door

Direct sell-side pressure from a major corporate entity exacerbated the macroeconomic headwinds. Strategy executed a massive treasury sale. The company offloaded 1,690 Bitcoin between August 3 and August 9 at an average price of US$64,262. The corporate entity successfully raised US$108.6 million to repurchase preferred stock. This transaction represents a small fraction of their total 840,447 Bitcoin holdings.

The timing proved disastrous for stability. Dumping over US$100 million worth of tokens into an illiquid order book inevitably crushes the price. This strategic shift rattles confidence because the community previously viewed this specific corporate holder as a permanent accumulator. The introduction of concentrated supply fundamentally alters the short-term dynamics. This action provides a concrete catalyst for breaching crucial support thresholds.

Ethereum also suffered significant underperformance. The second-largest network dropped over 3 per cent and broke below the psychologically vital US$1,900 threshold. This technical breakdown triggered a cascade of automated stop-loss orders. The derivatives space amplified this downward force dramatically. Total open interest actually rose 6.93 per cent. This metric indicates that speculators aggressively opened new short positions rather than simply closing existing ones.

Bitcoin liquidations surged 120.95 per cent in a single day. Exchanges wiped out over US$51.73 million in leveraged long bets. These forced closures create a vicious feedback loop. Exchanges liquidate over-leveraged long positions and automatically sell the underlying asset. This mechanical process pushes the price lower and triggers further liquidations. This mechanical unwind severely damages market structure and accelerates the downward trajectory.

Sentiment currently reflects deep caution. The CMC Fear and Greed Index sits firmly at 37. This reading indicates widespread fear among retail and institutional participants. Technical indicators confirm this bearish outlook across multiple timeframes. Bitcoin recently broke below its 50-day moving average of US$64,686. The asset also violated the critical 78.6 per cent Fibonacci retracement zone near US$64,105. The broader ecosystem simultaneously tests its pivot point at US$2.18T.

Allocators are now focusing intensely on the critical Fibonacci support zone at US$2.15T for total capitalisation. Algorithmic trading systems monitor these exact mathematical thresholds. Computers execute automated sell orders when prices breach these lines. This automated behaviour makes these mathematical thresholds self-fulfilling prophecies when breached.

Also Read: Bitcoin’s 73% correlation with gold forces investors to rethink crypto

The immediate future hinges entirely on upcoming macroeconomic publications and central bank decisions. Traders eagerly await the United States July Consumer Price Index report on August 12. This inflation metric will dictate short-term direction. If the numbers show cooling inflation, participants will price in a higher probability that the Federal Reserve will pause rate hikes at its September 16 meeting.

A pause in monetary tightening typically boosts risk instruments by preserving liquidity. Stubborn inflation metrics will force the central bank to maintain higher interest rates. Spot Bitcoin exchange-traded funds might provide a crucial counterbalance to this downward force. These funds attracted a net inflow of US$98.9 million last Friday. Sustained institutional buying through these regulated vehicles could eventually absorb the excess supply and stabilise the price action.

Technical analysis outlines two distinct scenarios. The base case involves holding the US$2.15T support threshold. Defending this line allows prices to consolidate and build a stronger foundation. Breaking this threshold risks a severe retest of the US$2.04T yearly low. Bitcoin faces a similar binary outcome. The premier cryptocurrency must hold its recent swing low near US$63,750. Successfully defending this line enables consolidation between US$64,100 and the US$65,400 resistance zone.

Failing to hold US$63,750 opens the floodgates for a rapid descent toward the US$62,000 major support area. The current downturn stems from disappointing economic metrics, targeted corporate selling, and severe technical breakdowns. Leverage unwinds always accelerate these moves and punish overconfident speculators. My analysis suggests the digital space must accept its new identity as a highly correlated risk instrument.

Survival requires strict risk management and a keen eye on traditional indicators. The path forward depends entirely on buyers defending critical support zones and the Federal Reserve accommodating risk instruments. Participants must closely monitor upcoming inflation prints and central bank communications. Only stabilising macroeconomic cues can provide a durable floor and restore confidence among hesitant allocators.

Market watchers must remain vigilant as these macro variables unfold. The transition from a niche speculative asset class to a deeply integrated component of the global financial system brings immense volatility. Traders must adapt their strategies to navigate this complex landscape successfully.

—

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 Is the US$63,750 line the only thing standing between Bitcoin and US$62,000? appeared first on e27.

Southeast Asia’s tech sector raised US$4.779B across 17 rounds in July 2026, according to Tracxn, the strongest month in the tracked 12-month period, up 25.53% from June and 180.9% year-on-year.

Two mega-rounds drove the surge: Kling AI’s US$2.8 billion raise and Ant International’s US$1.2 billion round, together accounting for roughly US$4 billion, or the bulk of the month’s total (both companies are Chinese-founded but now headquartered in Singapore). Without these two deals, July’s haul would have been far less dramatic.

The next-largest rounds were PixVerse (US$139 million) and dConstruct Robotics (US$125 million), followed by Whale, Ropedai, Rize, Tikva Allocell, Paypartners and Haup.

The stage-wise breakdown showed seven early-stage rounds, five seed and five late-stage deals, a relatively balanced pipeline despite capital concentration at the top. Active investors included Singtel Innov8, Lollapalooza Capital, Altara Ventures and Breakthrough Energy.

The numbers point to a two-speed funding environment: well-positioned companies in AI, fintech, robotics and climate tech can still command outsized cheques, while early-stage founders without clear traction face a tougher road. Investors remain selective even as headline momentum builds.

REGIONAL

NUS, OpenAI widen AI tie-up to cover all students, staff: NUS will give every student, faculty member and staff member access to ChatGPT Edu and Codex under an expanded OpenAI partnership, as a survey found 94% of Singapore university students already use AI weekly.

Vietnam’s VinSpace books SpaceX ride for 2027 satellite launch: VinSpace, part of Vingroup, signed its first launch contract with SpaceX to send Vietnamese-made nano-satellites into orbit via a Transporter rideshare mission in 2027.

Malaysia ranks third globally for AI use in wealth management: 85% of Malaysia’s affluent investors use AI for finance and investment decisions, per HSBC, trailing only India, though 58% still want AI paired with human expertise.

Vietnam fines Grab US$51,700 over consumer protection breaches: Vietnam’s competition regulator fined Grab Vietnam over failures to let users control data sharing and disclose influencer sponsorships, despite Grab’s 2025 revenue climbing 20% to US$3.37 billion.

Singapore AI adopters gain revenue, jobs, but profits lag: A Ministry of Trade and Industry study found AI-using Singapore firms saw revenue and employment gains, but no statistically significant profit boost within four years of adoption.

INTERVIEWS & FEATURES

Southeast Asia must move from connector to decision-maker: With US-China neutrality growing costlier to maintain, the region’s advantage lies in translation and adaptation, requiring heavier investment in home-grown research.

INTERNATIONAL

OpenAI completes US$7 billion employee share buyback: OpenAI bought back US$7 billion in employee shares at an US$852 billion valuation, a move seen as easing pressure for a near-term IPO.

Meta, TikTok face thousands of addiction suits after ruling: A US appeals court rejected platforms’ bid to dismiss thousands of addictive-design lawsuits via Section 230, allowing consolidated litigation to proceed.

Ant Group leads funding round for China’s Daimeng Robotics: Ant Group led a fresh funding round worth hundreds of millions of yuan into Daimeng Robotics, a Shenzhen tactile-sensing startup expanding into data infrastructure.

Bezos nears stake in Liverpool FC amid US buyout wave: Jeff Bezos is reportedly close to buying at least a 30% stake in Liverpool at a £1.35 billion valuation, joining a long list of American billionaires in the Premier League.

CYBERSECURITY

Crypto crime turns physical as ‘wrench attacks’ surge: Chainalysis reports over US$30 million stolen in violent crypto attacks globally in 2026 so far, with France, the US, Brazil and Thailand worst hit as home invasions and family-targeting both rise.

US$7B Philippine cyber modernisation sits wide open for SEA firms: The Philippines’ ₱430 billion defence modernisation drive faces a severe cybersecurity skills gap, leaving a rare, largely uncontested opening for regional systems integrators.

Bybit sues North Korea, Lazarus Group over US$1.5B hack: Bybit filed a US civil suit against North Korea, its intelligence agency and the Lazarus Group over last year’s record US$1.5 billion Ethereum theft, securing a court order freezing stolen assets.

China’s Kimi K3 model escapes sandbox during cyber test: Moonshot AI’s Kimi K3 broke out of an isolated test environment and searched GitHub for answers during a UK-run evaluation, the latest in a string of AI models evading containment.

Open-weight AI models close gap with frontier, safety lags: Advocates say open-weight models like GLM-5.2 helped Hugging Face defend against an OpenAI-model-driven breach, but critics warn wider access to near-frontier capability raises misuse risk.

SEMICONDUCTOR

Nvidia, Wall Street giants unveil US$500B AI chip financing plan: Nvidia partnered with Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR to mobilise over US$500 billion in third-party capital, treating AI compute as a bankable asset class.

Powertech pours US$400M into Singapore AI chip packaging plant: Taiwan’s Powertech Technology will take a 30% stake in a US$5.66 billion Broadcom joint venture building AI chip packaging capacity in Singapore, with AMD as first client.

Microsoft to unveil Maia 300 AI chip as early as September: Microsoft plans to publicly reveal its next-generation Maia 300 chip this autumn and is negotiating with TSMC for over 300,000 units by 2027.

Struggling AI hedge fund doubles down with US$400M chip bet: Situational Awareness, the AI-focused hedge fund whose assets nearly halved this year, invested a further US$400 million in stealth chip-maker Source Foundry, taking its total stake to US$500 million.

AI

Agentic AI’s next big market may be the back office: A Sunrate-Mastercard report projects B2B agentic payments to grow at a 335% five-year CAGR, far outpacing consumer transactions, as invoice processing and FX conversion become AI’s first quick wins.

THOUGHT LEADERSHIP

Southeast Asia’s ‘biggest market first’ expansion logic is dead: TikTok Shop’s Indonesia ban-and-pivot into Tokopedia, plus Jakarta’s platform fee cuts, show ranking markets by size no longer works for founders.

Vietnam’s ‘born global’ startups skip the home-market stage: Vietnamese founders increasingly build for international markets from day one rather than expanding abroad after domestic success, drawn by improved payments and cloud infrastructure.

Why optionality is Southeast Asia’s last real advantage: A hospitality founder argues the region’s refusal to pick a US or China bloc, plus Indonesia’s domestic-demand resilience, gives it hedge value increasingly scarce elsewhere.

Good ideas are everywhere, venture capital isn’t: Venture capital rewards ecosystems, not just ideas — Singapore’s institutional density and Silicon Valley’s recycled talent explain funding gaps between comparable startups.

Filipino virtual assistants deserve pay beyond ‘cheap talent’: A veteran VA argues clients conflate affordability with low value, urging businesses to compensate VAs for the complexity and specialised skills many now bring.

Market share is not power, control points are: Durable business advantage comes from owning unglamorous choke points — billing, identity, compliance, data custody — rather than customer volume.

How to tell a real AI marketing agency from a wrapper: Buyers should test agencies on platform ownership, transparency and experiment velocity, but human-in-the-loop oversight remains the factor separating growth from reputational damage.

The most dangerous place for a good idea is your head: Southeast Asian workers and SMEs sit on unclaimed inventions born from daily workarounds; AI can now lower the cost of testing a rough idea.

Bitcoin’s BIP-110 fork collapses within eight hours: An attempted Bitcoin protocol split drew just 2.53% miner support and stalled almost immediately, removing a source of uncertainty as institutional ETF inflows continue.

The post Ecosystem Roundup: SEA tech funding hits US$4.78B in July, led by mega rounds appeared first on e27.

Vietnam’s startup ecosystem has spent the past few years proving it can produce consumer apps, fintech platforms and edutech companies at regional scale. N2TP is attempting something less common, and arguably harder: building AI infrastructure for scientific research, where outputs are not measured in clicks or transactions, but in hypotheses, experiments, papers and patents.

The Hanoi-based company has raised seed funding from Touchstone Partners, the Vietnam-focused VC firm known for backing AI and deeptech companies such as Alpha Asimov, Eureka Robotics, Forte and Prep. The size of the round was not disclosed.

Also Read: AI infrastructure: The unsung hero of technological innovation

Founded in 2020, N2TP describes itself as an “AI Lab” focused on scientific research and intellectual property development in fields including AI infrastructure, biomedicine and biotechnology. Its core product is the N2TP AI4Science Platform, which aims to help research teams move beyond using AI merely to speed up isolated tasks and towards using it as part of the scientific process itself.

That distinction matters. In science, an AI-generated answer is not a discovery. A model may suggest a new molecule, pathway or biological relationship, but researchers still need to test whether the idea is logically sound, grounded in domain knowledge, experimentally feasible and reproducible. N2TP’s platform is designed to sit inside that loop: generating and assessing hypotheses, supporting simulation, helping design experiments, collecting new data and feeding results back into the system.

For Southeast Asia, where many research institutions and startups operate with tighter budgets than their counterparts in the US, Europe or China, this kind of infrastructure could be significant if it works at scale. The region has strong scientific talent but often lacks the same depth of capital, automated lab infrastructure and commercialisation pathways. AI-for-science tools could help narrow that gap, though they will not remove the need for serious laboratory validation.

From research bottleneck to repeatable loop

N2TP was founded by CEO Duong Thi Hong Nhung, a doctoral candidate at Hanoi University of Pharmacy who holds a master’s degree in pharmaceutical biochemistry; Chief Product Officer Do Ngoc Tuan, a computer science graduate of Goldsmiths, University of London; and Head of Research Operations Ho Hai Phong, who studied engineering at Kyushu University and finance at Waseda University in Japan.

The mix of pharmaceutical science, computer science and research operations reflects the problem N2TP is trying to solve. Scientific discovery is not slowed down only by a lack of ideas. It is slowed down by the work needed to turn an idea into something testable, then into evidence, then into a product, paper or patent.

N2TP says its AI4Science platform can help narrow the search space early by eliminating options that do not meet scientific or operational constraints. After a hypothesis has been validated through simulation, the system can support the next steps: experimental design, measurement and data collection. Those results can then be used to update the model and guide the next round of work.

In practical terms, this is less about replacing scientists than about building a tighter feedback loop between computation and experimentation. That is especially relevant in areas such as drug discovery and biotechnology, where teams may need to evaluate huge numbers of possible compounds, biological targets or experimental conditions before arriving at a viable path.

“We do not see AI as a tool to replace scientists. Scientists are still the ones who ask the questions, set the standards, oversee the process, interpret the results and bear responsibility for important decisions,” Nhung said. “What N2TP aims to build is infrastructure that more tightly connects hypothesis, reasoning, simulation and experimentation.”

Early output, but commercial questions remain

The company claims its platform has already improved research productivity. Over the past 12 months, N2TP says it has had 12 research papers accepted, presented or published at major global research conferences and forums, including ICML 2026, ACL 2026, UAI 2026, SIGMETRICS 2026, AAMAS 2026 and ISMB/ECCB 2025. It has also published three papers in Q1 journals: Scientific Reports, CPT: Pharmacometrics & Systems Pharmacology and Computers in Biology and Medicine.

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

In addition, N2TP has filed seven patent applications in Vietnam and internationally across foundational AI, biomedicine and biotechnology. The company says this pace is at least four times faster than its own output under a traditional research model.

Those numbers are useful markers, but they are not the whole story. In deeptech, publications and patents show capability, but commercial value depends on whether the underlying technology can be turned into defensible products, licensing revenue, partnerships or internal drug and biotech pipelines. Many AI-for-science companies globally have found that strong models are only one part of the equation; access to high-quality data, wet-lab validation and regulatory pathways can be just as decisive.

N2TP plans to use the new funding for two main areas: developing its patent portfolio in strategic technologies and completing the research loop through deeper integration with automated laboratory processes and equipment. The latter will be important. AI systems become more useful in science when they can learn from experimental results quickly and repeatedly, rather than relying only on existing datasets.

A crowded global field, a quieter regional one

N2TP is entering a global market that has attracted serious capital and talent. Google DeepMind’s AlphaFold changed expectations for AI in biology by predicting protein structures at scale, while Isomorphic Labs is applying similar capabilities to drug discovery. US-listed Recursion uses machine learning and large-scale biological datasets to build drug pipelines, while Hong Kong-founded Insilico Medicine has become one of Asia’s most visible AI drug discovery companies.

Compared with these players, N2TP is at an earlier stage and is building from Vietnam, where deeptech capital is still developing. Its advantage, if it can sustain one, may come from a focused team, lower R&D costs and the ability to build intellectual property around specific scientific workflows rather than compete head-on with global giants across every part of the AI biology stack.

Within Southeast Asia, the field remains comparatively thin. The region has produced healthtech, diagnostics and biotech startups, but fewer companies are building foundational AI infrastructure for scientific discovery. That gives N2TP room to define a category locally, but also means it may need to look beyond Vietnam early for partners, customers and validation.

Why Touchstone is betting on harder tech

For Touchstone Partners, the investment fits a broader push into Vietnam’s deeptech sector. Since launching in 2021, the firm has backed companies across AI, robotics, education, agriculture, healthcare and climate technology, including through initiatives such as the Net Zero Challenge.

The N2TP deal also reflects a growing belief among some Vietnamese investors that the country should build more than application-layer startups. While software products can scale quickly, foundational intellectual property in AI, semiconductors, biotech and advanced manufacturing is increasingly seen as important for national competitiveness.

“N2TP shows that Vietnamese researchers are fully capable of building core technology that meets international standards, even in as demanding a field as biomedicine,” said Ngo Thuy Ngoc Tu, Director of Touchstone Partners. “We believe that AI infrastructure for scientific research will be a critical piece of Vietnam’s technological development in the years ahead.”

Also Read: Razer and NUS launch Singapore AI lab to rethink how games respond to players

The hard part starts now. N2TP has early research output, a technical thesis and new venture backing. To become more than a promising lab, it will need to show that its platform can produce repeatable scientific and commercial outcomes.

For Vietnam’s startup ecosystem, the company’s progress will be watched not only as a funding story, but as a test of whether the country can build deep technology companies whose value lies in original research and defensible IP. That is a slower path than most startup playbooks allow, but it may be the one that matters most.

The post Touchstone backs Vietnam’s N2TP to build AI infrastructure for scientific discovery appeared first on e27.

Imagine two founders raising their first million dollars this year.

One is in San Francisco. She emails a former operator who exited a company in her space. He replies in two hours. They meet the next morning. The term sheet arrives within the week.

The other is in Singapore. He pitches a regional VC firm. He waits for the partner meeting. Then the investment committee. Then the second IC. Three months in, the firm passes. He starts over with another fund.

These two founders are building similar companies. They are facing similar markets. But they are raising capital on completely different playbooks. The first founder is operating on the new US model. The second is still on the old one.

The US version is coming to Southeast Asia. Within the next two to three years, the founders who understand this will have a structural advantage. The ones who do not will be running last decade’s race.

Here are the five shifts to track.

Shift one: The rise of the one-person fund

In US venture capital, the most influential investor in many deals is now a single person. No partners. No committee. No quarterly approval process. Just one investor making a call.

These are called solo GPs. They now make up more than half of all new fund managers globally. The most famous of them, Elad Gil, raised a billion-dollar fund on his own in 2024. Partners at top firms like Sequoia are leaving to do the same.

Why does this matter for a Southeast Asian founder? Because the speed advantage is dramatic. A solo GP can decide on a deal in days. A traditional VC firm takes months. When you are racing to ship a product, that gap is the difference between catching a market and missing it.

India is already seeing solo GPs rise. Southeast Asia is next.

Shift two: Operators are beating institutions for the best founders

Five years ago, the best founders in the US wanted Sequoia or Andreessen Horowitz on their cap table. The brand was the prize.

Today, many of those same founders are choosing someone different. They are choosing the operator who built a similar company ten years ago. The investor who knows the playbook because they wrote it themselves. The check writer who can pick up the phone and introduce them to their first ten customers.

Brand has not stopped mattering. But it has stopped being decisive.

What changed? AI made building faster. A small team can now ship a product, find customers, and hit revenue in months. Founders moving at that speed cannot afford an investor who moves at quarterly committee speed. They need someone who has been in the trenches and can answer the hard question on the same day.

Also Read: Connecting founders across Southeast Asia used to be the easy part of the job, and now it’s becoming the whole job

The Southeast Asian founders who win in 2026 will increasingly choose their investors the same way.

Shift three: The middle of the funding ladder is disappearing

For two decades, the path was simple. Raise seed. Then Series A. Then B. Then C. Each stage had its own investors, its own valuations, its own playbook.

That ladder is breaking.

At the bottom, solo GPs and operator angels are taking the early deals before the traditional firms can run their process. At the top, mega-funds are writing the giant cheques into AI companies. The middle, where most traditional partner-stage VCs lived, is becoming empty.

Southeast Asia is showing the same pattern. In the first quarter of 2026, regional startups raised US$2.81 billion. Sounds healthy. Look closer and the picture changes. That money was spread across just 98 deals, the lowest quarterly count in eight years. A handful of mega-rounds carried the entire quarter. Singapore alone absorbed over 90 per cent of the capital. The middle has thinned.

If you are a founder raising a Series A in Southeast Asia today, you may already be feeling this. The firms that used to be there are quieter. The deals that close are either small and fast at the bottom, or huge and concentrated at the top.

Shift four: Selling shares before IPO is becoming normal

US founders used to have one way to get personal liquidity. Wait for the IPO. That could take ten years. Sometimes longer.

A new path has opened. It is called the secondary market. Founders, early employees, and sometimes even VCs sell portions of their shares to other investors before the company exits. In 2024, this market hit US$160 billion in transaction volume globally. In 2025, it crossed US$210 billion.

For Southeast Asian founders, this matters because the IPO window here has been effectively closed for three years. Waiting for the public market to reopen is not a viable personal financial plan. The founders who learn how secondary liquidity works, and how to negotiate it into their later rounds, will have options that their peers do not.

Most Southeast Asian founders have never thought about this. Their global counterparts have.

Also Read: Founders’ playbook: What it really takes to scale beyond Series A

Shift five: The cheque has become the least valuable thing investors offer

Ask a US founder what they want from an investor in 2026. Capital will not be the first answer.

They will say distribution. Customer introductions. Hiring networks. Help with positioning. Strategic advice when the pivot fails or growth slows. The cheque is assumed. Everything around the cheque is the actual product.

This is the shift Southeast Asian founders are least prepared for. Most regional accelerators and VC firms still pitch themselves on the bundle of money, mentorship, and demo day access. The Y Combinator playbook from 2010.

In the US, that bundle has been taken apart. Founders evaluate investors on each capability separately. Money is a commodity. Everything else is differentiation.

The Southeast Asian founders who learn to evaluate investors this way are going to make very different decisions than the ones who do not.

What to do about it

None of these shifts will land in Southeast Asia in exactly the same way they did in the US. Capital structures here are different. Regulation is different. The culture of risk is different. But the directional reality is clear.

Three actions for founders raising in 2026 and 2027:

Start studying which Asian solo GPs and operator-investors are emerging. They are still few in number, but they are growing. Knowing them before the rest of the market does is the kind of asymmetric advantage that compounds.

Treat your cap table as a strategic asset. Every cheque carries non-financial implications. The investor who solves your distribution problem is worth twice as much as the investor who just adds a logo.

Understand secondary liquidity before you need it. The founders who walk into their Series B already knowing how to negotiate secondary terms will leave more value on the table than the ones who learn it under pressure.

The founders raising in the next two years will define the next decade of Southeast Asian technology companies. The ones who study the US shift early will be building on the new playbook. The rest will spend the decade catching up.

—

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 5 US venture capital shifts every Southeast Asian founder should be tracking right now appeared first on e27.

San Francisco is not short of AI conferences promising to reveal the future. Most deliver a parade of demos and a lot of vague optimism.

Upscale Conf, organised by Magnific (the Spanish company formerly known as Freepik), was different, not because it avoided the hype, but because the people on stage and in the hallway conversations kept circling back to something more interesting than the technology itself: what happens to human creativity when the cost of producing it collapses.

Also Read: Magnific bets on human‑led AI infra for marketing and film work

Over several conversations with a Singaporean digital artist, an HBO-trained movie director, and the CEO steering one of the world’s fastest-growing AI creative platforms, five ideas kept resurfacing. None of them are the ones you’d expect from a typical AI conference recap.

1. The real risk isn’t AI replacing creativity; it’s AI replacing depth

Wenhui Lim, the Singaporean artist behind niceaunties, has spent years building an entire speculative universe around the figure of the Southeast Asian “auntie” using AI image and video tools. Her warning to founders wasn’t about job losses or copyright. It was about shallowness.

“If you use AI purely to extract value or to scale output, you will hit a wall,” she told e27. “The metaverse is a good cautionary tale; it felt like an escape from physical, human experience, and ultimately people returned to what connects us. AI has more longevity because it can be used to go deeper into the human experience, not away from it. But that requires imagination, not just a roadmap.”

It’s a distinction worth sitting with, particularly for Southeast Asian startups racing to bolt generative features onto existing products. Volume is easy now. Meaning is not. Lim’s other quiet provocation — that she dislikes the term “AI artist” because “there is no such thing, there are artists working with AI”– is a useful filter for any founder currently rebranding themselves around a tool rather than a point of view.

2. Hybrid is the only honest answer, and the scarcity mindset is the real enemy

Noah Wagner, a film director who spent seven years at HBO before making an AI-themed thriller a decade before generative tools existed, made a case that cuts against both the AI-skeptic and AI-maximalist camps. His argument: nothing about storytelling fundamentals has changed, even as everything about production has.

Wagner is currently juggling three projects that sit at wildly different points on the AI spectrum, from a fully generative claymation series to a romance feature that uses AI only for background environments, never for the human performances at its centre. His advice to studios chasing efficiency was blunt.

“Don’t go into any AI endeavour with a scarcity mindset; don’t lead with ‘where are we saving money?’” he said. “Go in with an abundance mindset: how do we maximise what we’re doing creatively? The savings tend to follow. It could be five per cent on one project, 50 per cent on another. It genuinely depends on the problems you’re solving and the people involved. But that should never be the starting point.”

Also Read: How creativity, commerce and AI collide in mid-2026 marketing mix

For Southeast Asia’s under-resourced but fast-growing content industries, that reframing matters more than any specific tool. Wagner’s point about democratisation wasn’t abstract flattery, either: “The same way it’s been empowering for me, it’s going to be empowering for anyone who has a story to tell but doesn’t have millions of dollars to tell it.”

3. Southeast Asia isn’t the next market; it’s already the biggest one

If there was a genuine surprise buried in the conference, it was this: Magnific’s largest user base by country isn’t in the US or China. It’s India, followed by Brazil, with the US in third place. Indonesia and Thailand aren’t far behind.

Joaquín Cuenca, Magnific’s co-founder and CEO, was refreshingly unbothered by the usual anxiety about American or Chinese AI dominance.

“People don’t look at the label to see if a product comes from the US or not; they just use the product that they want to use,” he said. He also made an unexpected case for being a European company operating in Asian markets: “For enterprise customers, it’s a little bit like Switzerland. It’s not the US, it’s not China. They know that their data is going to remain private.”

This isn’t a minor footnote for a Southeast Asia-focused audience. It suggests the region isn’t waiting to be served by generative AI tools built elsewhere; it’s already one of the primary users shaping how those tools evolve, price sensitivity and all. Magnific’s entry price sits around US$6 to US$8 a month depending on the plan, deliberately low enough for the price-sensitive markets that built its original user base.

4. Isolation from Silicon Valley can be a structural advantage, not a handicap

Cuenca’s own founding story runs counter to the standard startup script. He built his first company in Cox, a town of a few thousand people in southern Spain, bootstrapped Freepik without raising a single round of venture capital, and credits that isolation for the company’s discipline.

“We were not native speakers. We are different from the average entrepreneur in San Francisco,” he said. “It gave us time to grow our uniqueness in the south of Spain, quite isolated. And eventually, we became a strong player in the stock industry by being different.” That same distance from Silicon Valley’s conventional wisdom, he argued, is what allowed Magnific to rethink its business “from scratch” when generative AI arrived, rather than inheriting assumptions built for a different era.

For founders operating far from the usual capital hubs, a familiar condition across much of Southeast Asia, this ought to be reassuring rather than discouraging.

5. The next competitive battle isn’t the prompt; it’s organisational memory

Perhaps the most consequential announcement of the week had nothing to do with flashier image generation. Magnific unveiled a trio of enterprise products — MCP, Flows, and Agents — designed to solve a much less glamorous problem: what happens when only two people on a 40-person marketing team actually know how to get good results out of AI, and everyone else is stuck bottlenecking around them.

Also Read: Is AI the end of originality or a new dawn for creativity?

“Access isn’t the same as building,” Cuenca said. “Building means your team can run it, not just you. It means the AI remembers your work, not just your last message.”

Omar Pera, Magnific’s CPO, framed the ambition more plainly still: “We’re much more interested in what humans can do when they have the right tools. Our goal has never been to replace creators. It’s to give them the power to create things that previously required larger teams, larger budgets, or more time.” For agencies juggling campaigns across Jakarta, Bangkok, Manila, and Ho Chi Minh City simultaneously, that shift — from clever prompting to shared, governable workflows — may end up mattering more than any single model upgrade.

Taken together, these five threads point to a conference that was less about marvelling at what AI can generate and more about the harder, less photogenic work of figuring out what it’s actually for. Southeast Asia, it turns out, isn’t just watching that conversation from the sidelines. It’s already in the room.

The post The scarcity mindset is killing creativity, not AI appeared first on e27.