Southeast Asia’s AI story is often told as a race to build models or launch chatbots. The more immediate opening may be smaller and more practical: software agents that reduce fraud losses, clear operational backlogs, and automate repetitive casework inside large local industries. That matters in a region where the digital economy reached US$263 billion in gross merchandise value in 2024, yet trust, labour constraints, and uneven software adoption still slow execution, according to Google, Temasek, and Bain.

The first real gap is not intelligence but operational drag

The strongest case for AI agents in Southeast Asia is not that companies suddenly need more ideas. It is that many still run important processes through spreadsheets, email chains, call-centre queues, manual reviews, and fragmented software.

That makes payments operations, scam detection, BPO workflows, and hospital administration better targets than general-purpose assistants. These are environments where tasks repeat at scale, errors are expensive, and buyers can measure whether a product actually reduces handling time, fraud leakage, or case backlog.

Google, Temasek, and Bain say AI adopters in the region are already realising returns within 12 months. That is a useful signal for founders: in Southeast Asia, the first winning agent startups are likely to sell cost reduction and process reliability before they sell autonomy.

Fraud and compliance may be the clearest wedge

If one category looks especially underbuilt, it is digital trust. Singapore’s annual scam brief put 2024 scam losses at about S$1.1 billion, while the same e-Conomy SEA 2024 research found that half of digital users had already fallen victim to online scams despite feeling confident they could detect them.

That creates room for agentic products that sit inside existing financial and commerce workflows: transaction monitoring, merchant-risk review, scam-pattern escalation, claims triage, customer outreach, and internal investigation support. These are not glamorous use cases, but they map directly to budgets because the cost of doing nothing is already visible.

Also Read: Hospitality needs to treat AI agents like a new channel, not a new feature

The region’s policy conversation also points in this direction. ASEAN’s expanded guidance on generative AI highlights deepfakes, inaccurate outputs, privacy, and malicious activity as material risks, which means products that improve verification, auditability, and safe human handoff may have a more credible path to adoption than agents that try to operate unchecked.

The Philippines shows why back-office agents matter

The second major opening sits in business-process work. Reuters reported that the Philippine IT-BPM industry was on track for US$38 billion in revenue and 1.82 million jobs in 2024, with an industry roadmap targeting up to US$59 billion by 2028. That is exactly the kind of large, process-heavy market where agents can create value without needing fully autonomous decision-making.

A credible startup here would not try to replace entire teams. It would target narrower pain points such as case summarisation, post-call compliance checks, workflow routing, multilingual knowledge retrieval, or quality assurance for voice and chat operations. In those settings, the product succeeds only if it works with existing CRMs, ticketing systems, and service-level agreements.

This is also why the market remains underexploited. Local buyers may be interested in AI, but deploying agents inside live customer operations requires strong integration, careful change management, and clear accountability when outputs are wrong. Even in Singapore, IMDA said only 14.5 per cent of SMEs adopted AI in 2024, up from 4.2 per cent a year earlier, which suggests willingness is rising faster than readiness.

Healthcare administration is a harder market but a real one

Healthcare is often discussed as a moonshot AI category, but the nearer opportunity in Southeast Asia is less about diagnosis than administration. Philips, citing WHO estimates, said the region could face a 6.9 million health-worker shortage by 2030, while surveyed professionals pointed to burnout from non-clinical work and concern about growing patient backlogs.

That makes scheduling, referrals, documentation, discharge coordination, and claims preparation more plausible starting points. An agent that reduces clerical load for nurses, front-desk teams, or care coordinators may be easier to validate and regulate than one that makes frontline clinical judgments.

Also Read: The one-person company was always possible. AI agents make it probable

The same constraint applies across the region: the best products will be the ones that respect low-trust environments. Founders need to assume patchy data, mixed-language workflows, limited technical staff, and buyers who want human oversight embedded from day one.

Early signals are real, but the bar is higher than hype suggests

There are signs that agent-native companies are starting to form. Tech in Asia reported that 35 Southeast Asian startups had raised at least US$1 million for agentic AI products by early 2026, while Singapore-based Level3AI announced a US$13 million seed round to expand enterprise voice and chat automation. Antler separately said it invested US$7.4 million across Southeast Asian AI startups in the first half of 2025, including seven companies through an AI residency in Singapore.

Still, these are early signals, not proof that the region has found its defining winners. The agent startups most likely to work in Southeast Asia will be narrow, local, and deeply operational: they will sell into industries with measurable pain, build around regulation instead of around it, and keep a human in the loop where trust is low.

For founders, investors, and ecosystem builders, the call is straightforward. Stop asking where the region’s AI equivalent of a universal assistant will come from, and start backing teams that can remove a costly hour, a missed payment flag, or a hospital admin bottleneck from a real workflow in Southeast Asia. That is where the gold rush is likely to become a business.

—

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

The post Southeast Asia’s AI agent opening is in messy workflows appeared first on e27.

Around a week before I started writing this, I saw a job post on social media that was everywhere. It got a lot of attention, so I clicked to see why. The requirements were:

• Minimum bachelor’s degree.
• More than 15 years of work experience.
• Under 25 years old.

There were plenty of other detailed requirements, too, all tied to the role itself.

My first thought was that it had to be a typo. But after a few minutes, I started thinking about it differently. What if it was not a mistake? What if the recruiter was genuinely looking for someone young, flexible, and still able to deliver senior-level output?

It sounds extreme, but plenty of companies are looking for someone like this.

Why SMEs create unrealistic roles

Most SMEs do not create impossible roles because they like setting impossible standards. They do it because they are under pressure. They need growth, speed, and solid execution, but they often cannot afford multiple hires or senior-heavy teams.

The work still needs to get done. Marketing, operations, customer support, hiring, and admin all have to keep moving. But instead of building a team with clear roles, they squeeze several responsibilities into one position. They look for someone who can think strategically, move fast, adapt easily, and work with very little support. Sometimes they are not hiring for one role at all. They are really asking one person to carry the weight of several jobs.

At the core, they are trying to solve a real business problem with limited resources. 

Why AI agents make this fantasy feel possible

This is where AI agents start to change the picture.

The idea of one person doing the work of several people no longer feels unrealistic. When one employee has AI tools that can draft emails, summarise documents, organise information, suggest ideas, and automate repetitive tasks, that person can suddenly work much faster.

That is a big part of why AI agents are so attractive, especially for lean businesses. They feel efficient, flexible, and scalable. They can help marketers come up with content ideas, summarise campaign results, and prepare reports. They can help recruiters screen CVs, draft outreach messages, and manage interview schedules. In short, they can take a lot of repetitive work off people’s plates.

Also Read: Hospitality needs to treat AI agents like a new channel, not a new feature

And this is not just hype anymore. McKinsey’s 2024 global survey found that 65 per cent of respondents said their organisations were already using generative AI regularly in at least one business function. That does not mean AI is replacing entire roles, but it does show that companies are starting to treat it as a practical tool for everyday work, not just an experiment.

From a business point of view, the appeal is clear. AI agents look like a way to increase output without immediately adding headcount.

The real barrier is not only technology, but trust

The hardest part of using AI agents is not just the technology itself. It is trust, judgment, and the discipline to use them well.

It is fairly easy to trust AI with first drafts, summaries, or simple admin work. It gets much harder when the task involves customer communication, compliance, or decisions that depend on context. That is usually where hesitation starts.

I do not fully trust AI agents all the time. Some tasks that seem simple enough to hand over still do not come back as well as I expect. Sometimes the output is wrong. Sometimes the judgment is off.

And businesses know those mistakes are not always minor. A weak internal summary can be fixed. A poor customer response can damage trust. A compliance mistake can create much bigger problems.

You can see this gap between experimentation and real confidence at the company level, too. McKinsey reported in 2025 that while almost all companies are investing in AI, only one per cent believe they are truly mature in how they use it. The same report says the biggest barrier to scaling AI is not employees, but leadership not moving the change forward fast enough. That matters because many companies talk about AI as if it is just a tool issue, when in reality it is also a management and workflow issue.

So the real question is not just whether AI agents can do the task. It is whether businesses are comfortable relying on them when accuracy, accountability, and judgment really matter.

What AI agents might quietly take away from workers, especially junior talent

This is the part of the conversation I think we should pay more attention to.

A lot of the work AI agents are starting to handle is repetitive, low-risk, and operational. But that is also where many junior employees used to learn. That first layer of work was never just simple work. It was often where people built judgment, picked up context, and grew professionally.

Think about a junior marketer who no longer has to come up with content ideas, write campaign summaries, or prepare monthly reports, or a junior recruiter who no longer needs to screen resumes, schedule interviews, or manage candidate documents. A lot of junior employees used to learn through exactly these kinds of tasks. By drafting, organising, following up, researching, and handling simpler work first, they built pattern recognition, confidence, and judgment over time. 

Also Read: AI agents and the new rules of business execution

If AI agents absorb too much of that layer, companies may become more efficient in the short term, but weaker in the long term. If junior employees do not get enough real experience, where will future managers and specialists come from?

That is why AI adoption should not only be framed as a productivity story. It should also be part of a capability-building conversation. A company can save time today and still create a talent problem for itself tomorrow.

What companies should do instead

This is why businesses should stop thinking about AI agents only as a replacement tool and start treating them as a work redesign tool.

The better question is not, “How many people can AI replace?” It is, “Which tasks should AI take on, which should stay human, and which should be done together?”

AI agents are useful for repetitive admin, first drafts, summaries, sorting information, and process support. Humans are still better at relationship-building, final decisions, ethical judgment, and high-stakes communication. In most cases, the strongest model is not AI alone or human alone, but humans working with AI.

AI can also raise the floor for less-experienced teams. It can help people perform more consistently in areas like drafting, research, reporting, coordination, and execution. But exceptional talent will still stand out. Judgment, creativity, systems thinking, leadership, and taste are still much harder to compress.

That matters because it changes what company-building can look like. A business in Singapore or Australia, for example, may not need one person to do everything locally. It may make more sense to keep the most strategic roles close to home, build human teams in emerging talent markets like Indonesia or the Philippines, and let those teams work alongside AI agents.

That model can work across many functions, from software development to finance and accounting to sales and marketing. The goal is not to replace people. It is to build a more realistic and scalable way of working around them.

For SMEs especially, this matters. The answer is not to keep searching for one magical employee who can do everything. The smarter move is to build better systems around real people.

Also Read: The one-person company was always possible. AI agents make it probable

The end of the all-in-one employee

AI agents will absolutely change the way companies work. They will raise expectations around speed, output, and how much one person can get done. But they should not become an excuse for unrealistic hiring or weak team design.

The real opportunity is not to expect humans to do even more just because AI exists. It is time to rethink how work is shared, where people add the most value, and how teams are built more realistically from the start.

AI agents will not replace workers. But they may replace the fantasy of the all-in-one employee that so many SMEs have been searching for. And honestly, that may not be a bad thing. That fantasy was never a sustainable way to build a business in the first place.

—

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

The post AI Agents and the end of the all-in-one employee appeared first on e27.

GoTo’s first quarterly profit is less a triumphant endpoint than a revealing pivot point for Southeast Asia’s tech narrative. After years of subsidy-fuelled expansion, the company has finally demonstrated that scale can translate into earnings, but only after becoming a markedly different business.

The headline numbers — positive net income, rising EBITDA, and strong cash flow — signal a long-awaited shift in discipline. Yet beneath them lies a more nuanced reality: profitability is being driven disproportionately by fintech, not by the ride-hailing and delivery engine that built GoTo’s brand. Payments and lending, with their superior monetisation dynamics, are now doing the heavy lifting, while on-demand services mature and grow more slowly.

This evolution raises a strategic question. Is GoTo still a super app, or increasingly a fintech platform with logistics attached? The answer matters because fintech brings both higher margins and higher risks, especially around credit quality, where disclosures remain thin.

Equally notable is what enabled this milestone: fewer incentives, tighter pricing, and a leaner corporate structure. Profit has arrived not through expansion, but through restraint.

For investors, the signal is clear. The era of growth at any cost is over. The next test is whether this new, narrower GoTo can sustain profitability without sacrificing relevance in an intensely competitive market.

REGIONAL

GoTo’s first quarterly net profit driven by fintech arm: Indonesia’s biggest digital ecosystem posted a first-quarter net profit of US$10.2M, reversing a US$21.8M loss a year earlier, with fintech revenue surging 58% to US$113.6M, now carrying the bulk of the investment case as on-demand growth slows.

eFishery founder handed nine-year jail sentence: Gibran Huzaifah was convicted of corporate deception, embezzlement, and money laundering after eFishery’s alleged revenue manipulation erased US$300M in investor value, rattling confidence in private-market due diligence across Southeast Asia.

TrueMoney Philippines winds down after a decade: The digital payments and remittance firm is ceasing all commercial operations in the country, having served over 5M customers and 20,000 SMEs across payments, lending, and insurance since its launch.

Indonesia mandates suspension reporting for under-16 accounts: Platforms must now publicly disclose how many accounts belonging to under-16 users they suspend, with rules affecting 70M children, though digital rights groups warn age checks expose sensitive data and are easily bypassed.

TikTok Go by Tokopedia targets Indonesia’s F&B market: Launched on April 29, the local services feature bridges content and offline visits via videos, live streams, and location-based discovery, as daily dine-in merchant orders reportedly rose more than 20 times in 2025.

Roblox moves to comply with Indonesia’s under-16 curbs: The US gaming platform has begun rolling out compliance measures following Jakarta’s tighter online rules for minors, though specific steps were not disclosed by the country’s communications minister.

Malaysia’s Qarbotech wins Grand Prix at SusHi Tech 2026: The agritech startup beat 820 applicants from 60 countries to claim JPY10M (US$62,000) for its nanocarbon photosynthesis-boosting agent, signalling growing global resonance for tropical-climate agritech solutions.

Cata raises US$5.3M to democratise F&B app tech: The Singapore-based consumer app platform closed an oversubscribed seed round led by Portage, enabling independent F&B and retail operators to launch branded apps with loyalty, payments, and CRM within days, with Germany as its first international market.

Cube raises US$3.7M to solve e-commerce data chaos: The Bangkok-based market intelligence startup secured Series A funding led by Betatron Venture Group to expand AI-enabled product tagging into North Asia and Latin America, targeting brands flying blind on fragmented digital shelf data.

FORMAS.AI closes US$3.98M pre-seed for AEC design platform: Led by Vertex Ventures Southeast Asia & India, the oversubscribed round will fund a platform orchestrating 60-plus AI models for architects, with users across 135 countries already generating over 500,000 designs organically since November 2025.

SiamDL raises US$7.8M to expand AI lending in Thailand: The Bangkok-based lender secured an oversubscribed Series A from international investors to scale its proprietary AI credit-scoring system, targeting underserved consumers and micro-entrepreneurs with thin financial records through its Bank of Thailand-licensed apps.

Philippines talent gap threatens strategic sector growth: A Monroe Consulting report warns that transformation is outpacing talent readiness across energy, finance, and tech, with above-market salary increases of 7–10% projected for automation, cybersecurity, and AI engineering roles.

Pandai’s low-cost growth playbook earns LSE 100x Impact spot: The Malaysian edutech startup scaled to over one million users with near-zero paid marketing and a 94% monthly retention rate, landing in the LSE initiative that identifies organisations with potential to improve one billion lives.

---

INTERVIEWS & FEATURES

Netbank CEO: We want to be the full meal, not just ingredients: CEO Gus Poston explains how owning the banking ledger rather than wrapping legacy infrastructure lets Netbank serve fintechs at startup speed, with FY2025 revenue up 88% driven primarily by QR.Ph payment volume growth.

SEON CEO: AI exposes weak risk operations, not fix them: Tamas Kadar warns that automating before defining decision ownership is the industry’s biggest mistake, with fragmented fraud and AML systems creating faster confusion rather than clarity across Southeast Asian markets.

---

INTERNATIONAL

Anthropic eyes US$900B valuation in new funding round: The Claude maker is in talks with investors for a raise that would top OpenAI’s valuation, as its annualised revenue reportedly hit US$30B driven largely by demand for Claude Code, though no term sheet has been signed.

Parallel Web Systems hits US$2B valuation after back-to-back raises: Founded by former Twitter CEO Parag Agrawal, the AI search and research API startup raised a US$100M Series B led by Sequoia just five months after its Series A, bringing total funding to US$230M with over 100,000 developers on the platform.

SoftBank plans US listing of AI robotics unit Roze at US$100B: Japan’s investment giant is preparing to spin off an AI and robotics company focused on data centre construction, potentially targeting a public debut as early as 2026 at a US$100B valuation.

Ant International bets on AI commerce infrastructure at scale: The Alibaba affiliate’s payments network now links 150M merchants with 2B consumer accounts across 220 markets, handling 20M daily transactions, while Alipay’s new tool enables merchants to accept payments made by AI agents.

US tech giants set to outspend China 7:1 on AI infrastructure: American hyperscalers led by Google, Microsoft, Meta, and Amazon are projected to spend over US$700B on AI infrastructure this year, versus an estimated US$105B by Chinese cloud providers constrained by chip export curbs.

Crypto equities plunge as big tech earnings impress markets: Macro pressures from surging oil prices and a hawkish Fed hammered crypto-linked equities (Robinhood fell 14%, Coinbase 6-8%) even as Bitcoin held near US$76,000, exposing how differently macro cycles transmit through digital asset layers.

---

CYBERSECURITY

After the breach: How companies respond now defines trust: A Penta report draws on 18 months of stakeholder sentiment data to argue that rapid transparency and visible leadership — not technical containment alone — now determine reputational recovery, with retail carrying the sharpest negative sentiment at minus 77.

SEA’s cyber market races toward US$10B fortress by 2030: With cyberattacks up 85% year-on-year in 2024 and the regional market valued at US$2.8B growing at 28.5% CAGR, Singapore-based Darktrace, Senzing, and CyberArk command 45% market share as detection times collapse from 21 days to 47 minutes.

SEA’s AI-fuelled cyber war: Six defining trends for 2026: As Southeast Asia’s digital economy surpasses US$1T, deepfake phishing attacks spike 150%, quantum-resistant cryptography becomes a regulatory mandate, and ransomware groups hit 40% more Indonesian SMEs, forcing startups and governments into agile defence.

US$1.5B crypto hack exposes exchange security gaps: The largest digital theft ever sent Bitcoin below US$80,000 and swung sentiment from extreme greed to fear overnight, calling for multi-layered technical infrastructure, human-centric protocols, and transparent asset management as the new security standard.

APJ finance faces 3.7B attacks as ransomware victims surge 204%: Akamai’s State of the Internet report found 92.3% of APJ finance sector attacks targeted banks, while zero-day and one-day vulnerabilities drove a 204% jump in ransomware victims, making zero-trust architectures and microsegmentation critical defences.

Diverse IT teams are a cybersecurity weapon, not just good HR: With women comprising only 34-40% of Southeast Asia’s tech workforce, organisations with homogenous teams miss region-specific attack patterns; diverse cultural and technical backgrounds improve threat detection, social engineering recognition, and compliance across borders.

Connected cars are becoming prime targets for cybercriminals: From the 2015 Jeep Cherokee remote hack to Tesla’s 2020 keyless entry vulnerability, smart vehicles face growing risks as decentralised security models like DePIN and Soarchain’s blockchain infrastructure emerge as more resilient alternatives to centralised systems.

---

SEMICONDUCTOR

Samsung Q1 profit surges 8x to US$38.9B on AI chip demand: Record-breaking memory chip revenue driven by AI server demand and tight supply pushed first-quarter operating profit to 57.2 trillion won,  surpassing Samsung’s entire full-year 2025 earnings, with strong second-half server memory demand expected as hyperscalers expand AI capacity.

TSMC exits Arm Holdings with US$231M share sale: Taiwan’s chip giant has fully divested its Arm stake acquired at IPO for US$51 per share, selling remaining shares at US$207.65 each through TSMC Partners, adding US$174M to retained earnings and completing a profitable exit from the British chip designer.

---

AI

Google DeepMind CEO expects AGI arrival by 2030: Demis Hassabis says the 20-year mission to build AGI is precisely on track, urging business leaders to treat the timeline as operational reality, and crediting DeepMind’s early success to pairing deep learning with reinforcement learning ahead of rising GPU power.

SEA’s AI agent opportunity lives in messy workflows: With the region’s digital economy at US$263B GMV, the clearest wins for AI agents are in fraud detection, BPO workflows, and healthcare admin — where operational drag, not a lack of intelligence, is the real bottleneck startups should target first.

How AI agents are quietly rewriting the growth marketing playbook: An always-on AI agent that flagged 18% of monthly ad spend burning with near-zero conversions illustrates how agents amplify human intelligence in marketing — but strategy, cross-functional trade-offs, and accountability still require named human owners, not autonomous loops.

AI agents and the end of the all-in-one employee fantasy: SMEs that have long searched for one person to do everything are finding that AI agents can carry much of the operational load, but the real risk is hollowing out junior talent pipelines before future managers have a chance to build judgment through real work.

Why startup founders shouldn’t trust AI to replace a PR team: A founder’s six-month experiment replacing his PR agency with AI tools produced flawless-looking outreach lists full of outdated contacts and generic thought leadership pieces, revealing that AI accelerates volume but cannot replicate the relational judgment of an experienced communicator.

---

THOUGHT LEADERSHIP

ESG tech’s next frontier is defensibility, not measurement: As sustainability disclosures are increasingly held to financial reporting standards, platforms that cannot explain data lineage, who touched a number, when, and why, will fail enterprise assurance tests, splitting the market between workflow tools and auditability-first infrastructure.

APAC founders must treat communication as a leadership skill: In a region where social media backlash goes regional within hours and regulations shift without warning, crisis-ready startups need pre-approved messaging templates, a single source of truth, and media relationships built before they are needed, not after a crisis hits.

The digital transformation lie SMEs have been sold: Most SMEs are drowning in disconnected tools rather than transformed by them, and the hidden tax of bad software — duplicate work, slow invoicing, blind spots — costs real margin daily; the winning platforms will be those that reduce friction before they sell a solution.

Accessibility in business is an ROI driver, not just compliance: With 1 in 4 US adults living with a disability and digital accessibility lawsuits rising annually, businesses that embed inclusive design into websites, documents, and physical spaces improve SEO, reduce legal risk, and expand market reach for all users.

Why exhibition leads fail and how to build a pipeline that converts: Most organisations invest heavily in booth presence but neglect the structured post-event follow-up process where momentum is actually won or lost: clear ownership, contextual data capture, and defined timelines matter more than visitor counts.

When collaboration systems break down in tech-driven workplaces: With 69% of APAC organisations on hybrid models and 6 in 10 employees reporting burnout, leaders must treat collaboration as intentional infrastructure, not a set of perks, using interactive digital tools to restore equity, alignment, and engagement across distributed teams.

Why eSIM is becoming as important as your passport: As global business travel spending grows toward US$1.57T in 2025, eSIM adoption is accelerating because connectivity must now be ready at the moment of arrival, not sorted out after landing, with travel eSIM users forecast to grow from 40M in 2024 to 215M by 2028.

The post Ecosystem Roundup: GoTo turns profitable, but the story has changed appeared first on e27.

The first wave of ESG tech won attention by making sustainability data easier to collect, organise, and present. That was enough when most firms were still trying to get a report out the door, stand up a dashboard, or show that they were at least taking measurement seriously. That phase is ending. There is a clear weight on sustainability-related disclosures being comparable, verifiable, timely, and understandable, while also requiring connected information between sustainability disclosures and financial reporting.

That changes the commercial logic of the software category. The enterprise buyer is no longer asking only whether a platform can calculate a number. Increasingly, the buyer needs to know whether that number can survive challenges from finance, internal audit, external assurance, regulators, and the board. This is the integrity gap in ESG tech. Many platforms can produce metrics. Far fewer can explain, with discipline, where each input came from, what happened to it in transit, which assumptions touched it, and why the final output should still be trusted. That is not a reporting nicety. It is becoming a buying criterion.

The market has moved from measurement to defensibility

The most important shift is not technical. It is institutional.

Sustainability information is no longer treated as a side narrative. It is treated as part of general-purpose financial reporting and explicitly says the information should be provided in a way that helps users understand connections across governance, strategy, risk management, metrics, targets, and the related financial statements. It also defines verifiability in practical terms, saying information is more useful when it can be corroborated directly or through the inputs used to derive it. That is a higher bar than modern ESG software was originally built for.

Why lineage matters more than teams realise

Data lineage sounds technical, but its real value is managerial. Data provenance in terms equivalent to a chain of custody, covering the generation, transmission, and storage of information in a way that can trace origin. That is exactly the language enterprise climate tech should now be borrowing. For a large company, especially in sectors like oil and gas, utilities, manufacturing, and heavy industry, a reported climate metric rarely comes from one neat system. It is assembled across operational data, finance systems, procurement records, supplier inputs, certificates, spreadsheets, manual adjustments, and judgment calls. If that chain is unclear, the number may still be useful for internal direction, but it becomes harder to defend as enterprise-grade reporting.

Also Read: ESG as strategic value: Why Asian boards must move beyond disclosure

This is not abstract. The core reporting principles of many protocols say transparency depends on a clear audit trail, references to methodologies and data sources, and information being recorded and analysed in a way that allows internal reviewers and external verifiers to attest to credibility. It also states that companies are responsible for the existence, quality, and retention of documentation so as to create an audit trail of how the inventory was compiled. It goes further and says an audit trail or similar mechanism is needed to show that no other entity is claiming the same environmental attributes. In other words, lineage is not an administrative garnish. It sits inside the claim itself.

The uncomfortable part for vendors is that dashboards are no longer enough

Many climate tech products were shaped by the priorities of an earlier market. They were built to speed up collection, improve completion rates, standardise templates, and give sustainability teams a cleaner operating interface. None of that is trivial, but it is no longer where strategic differentiation lives.

The next enterprise question is harsher. Can the platform preserve a defendable history of the number, not just its latest version? Can it show which emission factor was used at the time, who overrode a value, which source file was replaced, when a boundary changed, and whether prior periods were restated consistently? The emphasis is on connected information, consistent data and assumptions, and verifiability means the product has to support coherence across narrative, metrics, and financial context. A tool that calculates well but forgets its own reasoning is going to look weaker with every reporting cycle.

Auditability is becoming a commercial feature, not a compliance burden

This is the part many founders and product leaders still underplay. Auditability is often framed as a back-office requirement that slows the user down. In enterprise climate tech, it is becoming part of the product’s commercial value.

The reason is simple. Buyers are now under pressure to prove consistency between sustainability reporting and financial reporting, and assurance helps ensure that connectivity and consistency. That means a platform with weak controls around lineage, documentation, and historical traceability does not just create future compliance pain. It creates present-day buying friction. The product may look strong in a pilot, yet fail the moment the CFO, controller, assurance provider, or procurement team asks how the evidence trail is preserved.

Also Read: Why investors and customers are betting on ESG-aligned startups

That is why I suspect the climate tech category is about to split into two groups. One group will remain workflow-heavy and presentation-strong. The other will start behaving more like financial infrastructure, with native traceability, stronger version control, clearer source attribution, and a deeper understanding of what enterprise assurance actually demands. The second group is where the long-term value will sit. That is an inference, but it is a grounded one given where reporting standards and assurance expectations are heading.

What the next winners will build

The winning products will not simply promise better data quality. They will make integrity visible.

That means preserving source-level provenance rather than flattening everything into a final table. It means maintaining version history for methods, factors, boundaries, and mappings. It means showing not just the result, but the route to the result. It means keeping enough context so that an internal reviewer, auditor, or senior executive can understand why a number changed and whether the change was operational, methodological, or clerical. These are not exotic product choices. They are the practical extension of what the protocols already ask for in audit trails and documentation, and what is asked for in verifiability and connected information.

There is also a strategic opening here for firms that understand complex industrial reporting. Oil and gas is a good example. Many of the most material climate numbers in this sector are assembled across assets, partners, market instruments, engineering assumptions, and operational systems. That makes the integrity challenge harder, but it also means the value of strong lineage is more visible. In sectors where climate reporting is closely tied to capital allocation, operational performance, and external scrutiny, the software that best preserves evidence will start to look more useful than the software that merely looks modern. That is where differentiation becomes real.

Final thought

The climate tech market is maturing out of its measurement phase. The next contest will be about defensibility.

In the years ahead, the most valuable platforms will not be the ones that produce the smoothest ESG narrative. They will be the ones that can help an enterprise answer a much tougher question with confidence: where did this number come from, what changed it, and can we prove it. Once that becomes the standard buying conversation, data lineage and auditability stop being technical extras. They become the product.

—

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

The post The integrity gap in ESG tech: Why defensibility is the next frontier appeared first on e27.

For years, content marketing was something many founders knew they should be doing but rarely prioritised early in their company journey. Not because writing was difficult. But because publishing consistently required a workflow: topic selection, search validation, editing, formatting, distribution planning and platform adaptation.

All of this added up to something that looked less like a single task and more like a small team function, often requiring at least a marketing manager and supporting execution capacity.

AI agents are changing that timeline.

They are not simply helping entrepreneurs write faster. They are making it possible to start publishing earlier, before dedicated marketing teams exist.

That shift is already changing how startups build visibility.

Content marketing used to depend on coordination

A single article rarely exists in isolation.

Even a straightforward piece often involves:

• validating whether the topic matters
• structuring the argument clearly
• aligning tone with audience expectations
• preparing metadata
• formatting inside a CMS
• adapting captions across platforms
• deciding when and where to distribute it

None of these steps individually are complex.

Also Read: Malaysia’s Qarbotech takes top honour at SusHi Tech 2026 global pitch contest

Together, they create friction, especially for founders without a digital marketing background. Even my tech-savvy husband asked what a focus keyword or meta description was when he started planning content for his own website. That moment reminded me how invisible the mechanics of content marketing are until someone has to run them personally.

AI agents reduce that friction across the workflow, not just by drafting articles. As such, consistent publishing is now possible earlier in a company’s lifecycle than before.

Agents change execution capacity, not strategy responsibility

Many teams still treat AI as a writing shortcut. But agents are more useful as workflow collaborators.

Instead of asking whether AI can write a post, founders can now test whether a post is worth writing at all.

Agents can help:

• compare positioning angles
• explore topic clusters instead of isolated ideas
• adapt content across multiple platforms
• maintain tone consistency
• shorten experimentation cycles

This shifts content marketing from production-heavy work toward decision-heavy work.

Also Read: When collaboration systems break down in tech-driven workplaces and how to fix them

Earlier publishing changes startup visibility dynamics

In Southeast Asia especially, many startups operate without dedicated marketing teams in their early stages.

At the same time, online visibility shapes:

• investor discovery
• hiring credibility
• partnership opportunities
• category positioning

Previously, consistent publishing usually followed scale. Now, it can happen earlier.

That creates a different starting point for how entrepreneurs shape their narrative. Instead of waiting until a marketing function exists, companies can begin building presence while they are still defining their category.

Why this matters for how founders build early teams

There is another shift happening quietly alongside this. Many early-stage founders still prioritise hiring sales before marketing.

The logic is understandable. Revenue feels urgent. Pipeline feels measurable. Sales conversations feel closer to outcomes.

But without positioning, messaging and content, sales teams often end up building their own materials as they go. That slows conversations instead of accelerating them.

Also Read: When AI agents start deciding, what happens to human judgment?

I saw this directly in a previous startup I worked in, where the sales team visibly relaxed once someone finally owned messaging and content. They immediately shared with me a long list of materials they needed in order to sell the company’s products and services with credibility.

Traditionally, the alternative was hiring a marketing manager earlier than founders felt comfortable doing. AI agents are changing that trade-off.

Founders can now support early sales activity with structured messaging, lightweight editorial presence and consistent narrative positioning even before a full marketing function exists. This creates a more balanced setup: sales teams still focus on conversations, while founders establish the materials and context those conversations depend on.

At the same time, it raises expectations for founders and early operators, including sales teams. When agents can support parts of the marketing workflow, people can no longer rely entirely on role boundaries. Early teams increasingly need to work across positioning, messaging and execution layers rather than waiting for a dedicated function to exist.

And despite the technical framing around AI, prompt engineering still behaves more like an editorial craft than a precise science.

Agents can accelerate output. They cannot decide what your company should stand for.

That part remains human.

AI agents make editorial testing cheaper

One of the less obvious effects of this shift is how it changes experimentation.

Founders can now explore multiple editorial directions before committing time to full production.

Also Read: Crypto plunges, big tech earnings are strong. So why are markets nervous?

For example, when evaluating whether an announcement or industry development deserves coverage, it is possible to quickly test whether it works better as:

• a search-driven article
• a founder-perspective reflection
• a newsletter insight
• a platform-specific short-form post

Previously, testing those options required drafting each version separately or consulting a PR agency.

Now the decision can happen earlier in the process. That reduces the cost of experimentation, and lower experimentation cost usually leads to better positioning decisions.

Agents expose weak content strategy faster

There is understandable concern that AI-assisted publishing will increase generic content online. That risk exists. But weak positioning did not begin with AI.

Agents simply make it easier to replicate surface-level strategies across the market. If content depends entirely on summarising trends or repeating competitor narratives, the advantage disappears quickly once everyone has access to similar workflows.

What remains difficult to replicate is interpretation.

Editorial judgement still determines:

• whether a topic matters
• how it should be framed
• who it is for
• why it deserves attention now

Agents assist execution. They do not replace positioning.

Also Read: When collaboration systems break down in tech-driven workplaces and how to fix them

Founder perspective becomes more valuable, not less

Another common concern is that AI-assisted workflows reduce authenticity. In practice, the opposite may be happening.

As production becomes easier, differentiation shifts toward perspective. Readers increasingly respond to lived experience, practical lessons, framing decisions and clearly explained trade-offs.

Agents are strong at structure. They are weaker at conviction.

That makes founder-led thinking more visible inside content strategies rather than less relevant.

The real risk is outsourcing judgement

There is a genuine risk in agent-supported workflows. It is not automation replacing marketers.

It is founders allowing agents to decide what should be published.

Agents should reduce execution effort. They should not replace editorial ownership.

If positioning starts sounding interchangeable across companies using the same tools, the issue is usually strategy rather than technology.

The founders who benefit most from agents are the ones who remain intentional about what their content represents.

Content marketing is becoming earlier and more strategic

AI agents are already changing:

• who can publish consistently
• when founders can begin shaping visibility
• how quickly ideas can be tested
• what a marketing team needs to look like

But they are also raising expectations.

If everyone can publish faster, the advantage shifts toward people who know what is worth saying in the first place.

And in the interest of transparency, and perhaps slightly proving the point… yes, this article was written with the help of an AI agent too.

—

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

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

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

The post AI agents didn’t change how I write, they changed when I could start publishing appeared first on e27.

The first wave of large language models won attention by replicating personal work. They wrote emails, summarised documents, generated images and videos, drafted presentations, and produced usable code. That alone was enough to reshape how many people think about productivity.

But the more consequential shift is now underway. AI is beginning to move beyond individual outputs and into business processes — the kind of work that has required a team, a set of procedures, and institutional knowledge that no single person holds alone. That is a fundamentally different challenge.

A personal workflow usually lives inside one person’s head and one person’s screen. A business process lives across a team. It depends on handoffs, approvals, coordination across systems, data from multiple sources, and rules for dealing with exceptions. What looks like a single task is often a web of hidden coordination.

As a venture capital investor, I have recently met founders working on exactly this problem — building agents for e-commerce operations, family office mid- and back-office work, and insurance distribution workflows. The first is trying to replicate much of the e-commerce COO function, from product selection and creative design to merchandising, operations, and logistics coordination. The second focuses on family office workflows such as capital call management, treasury handling of idle cash, reporting, analytics, and forecasting. The third is tackling insurance distribution, from lead qualification and product comparison to documentation, onboarding, and follow-up.

None of these is tools for isolated tasks. They are attempts to codify work that has always depended on coordinated teams — and that, until recently, was considered too messy and too human to automate.

Also Read: When collaboration systems break down in tech-driven workplaces and how to fix them

In the first phase of generative AI, the benchmark was usually the quality of the output. Did the model produce a convincing sales pitch, a striking design, a useful analysis, or working code? Once AI enters a business process, the measure of success changes. The real question becomes whether the system can move work forward reliably, repeatedly, and within the right constraints.

That is where the challenge becomes more organisational than technical. It is one thing to know your processes well enough to run them. It is another thing to describe those processes with enough clarity, consistency, and structure for software to execute them repeatedly. The gap is not necessarily in understanding. It is codification.

A process may exist in many places at once: in a standard operating procedure (SOP); in a manager’s judgment; in a spreadsheet passed around by email; in a series of unwritten escalation habits; or in the head of an experienced employee. Humans can work across that ambiguity because they improvise, ask around, and handle exceptions informally. AI systems are far less forgiving. They need the process to be legible.

This helps explain why so much enterprise software has historically disappointed employees. Many internal knowledge bases are hard to search, dry to read, and detached from the immediate task. Workflow systems often digitise the container of work rather than the work itself.

A good example is the office automation (OA) system common in many Chinese enterprises and state-owned enterprises. In principle, these systems were designed to digitise approvals, document flows, announcements, and internal coordination. In practice, they often became digital wrappers around slow, manual, and bureaucratic routines. The interface changed, but the burden did not. Employees still had to chase approvals, assemble context, and push work forward by hand. The process looked digital on the surface while remaining stubbornly manual underneath.

Also Read: When AI agents start deciding, what happens to human judgment?

One recent McKinsey report on the future of AI in insurance offers a useful glimpse of what automation actually looks like inside a complex workflow. Rather than describing agentic AI as a single smart chatbot, McKinsey breaks the process into a set of specialised roles: one agent gathers and clarifies information, another profiles risk, another structures pricing and product options, another checks compliance and fairness, another decides whether a case can be approved or escalated, and another learns from feedback over time.

A North American insurer even used agentic processes to uncover “implicit judgments” that experienced underwriters had relied on for years and codify them into new rules and protocols. McKinsey notes that this kind of embedded expertise — once invisible, now formalised — could become a central part of a firm’s intellectual property. In other words, the act of making a process legible enough for AI may itself create something valuable that the organisation never knew it owned.

That is the crux of agentic work. Replicating a business process is not simply a harder version of writing assistance. It requires structure: a defined task, authoritative data, decision rules, guardrails, confidence thresholds, and clear points for human intervention when reality diverges from the ideal flow. These are not merely model problems. They are organisational problems.

AI agents raise the possibility of turning organisational knowledge from reference material into executable behaviour. The real test is not whether a firm has documented its processes, but whether it has described them with enough programmatic rigour that software can carry them out repeatedly. The companies that get this right may find that codifying their workflows is not just an IT project. It is a way of discovering — and preserving — how the business actually works.

That, to me, is the real significance of AI agents. The age of agentic work may not be defined by whether machines can sound human. It may be defined by whether organisations can become legible enough for machines to work inside them.

—

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 The rise of agentic work: Can AI replicate a team, not just a person? appeared first on e27.

The post Stocks hit record highs while US$300M in crypto longs get liquidated: What’s next? appeared first on e27.

Indonesia has handed eFishery founder Gibran Huzaifah a nine-year prison sentence, bringing a brutal legal end to a fraud scandal that vaporised roughly US$300 million in investor value and shattered one of Southeast Asia’s most celebrated startup narratives.

The ruling, delivered by the Bandung District Court on April 29 and first reported by Bloomberg, marks a rare moment in the region’s tech industry: a once lionised founder of a unicorn-scale startup being convicted in a criminal case tied to corporate deception, embezzlement and money laundering. For Indonesia’s startup ecosystem, it is not just a courtroom verdict. It is a public reckoning.

Also Read: “There’s no excuse”: Aqua-Spark calls out eFishery’s deception

Huzaifah, the former CEO of eFishery, was also ordered to pay a fine of roughly US$60,000. Prosecutors had earlier sought a 10-year prison term. He now has seven days to appeal.

That still leaves the same uncomfortable conclusion for VC firms, founders and boards across Southeast Asia: one of the region’s flagship agritech companies did not simply fail. It unravelled into a governance disaster on a scale large enough to rattle global investors, damage confidence in private-market due diligence and expose how easily startup mythology can outrun financial reality.

From agritech darling to courtroom collapse

For years, eFishery was held up as proof that Southeast Asia could produce category-defining startups outside the usual ride-hailing, fintech and e-commerce script. The company, which supplied smart feeders and related services to fish and shrimp farmers in Indonesia, was once valued at more than US$1 billion. It had the right ingredients for a VC success story: a large domestic market, a sector with real-world impact, and a founder selling not just software but national relevance.

That story has now ended in disgrace.

According to Bloomberg, a panel of judges found Huzaifah guilty after a case centred on allegations that eFishery’s financial statements had been manipulated over several years. The company’s collapse followed a board investigation that raised concerns about inflated revenue and profits. What followed was a slow-motion demolition of one of Indonesia’s most prominent tech brands.

The investor list makes the fallout especially painful. Backers caught in the wreckage include Temasek, SoftBank Group, Peak XV, and 42XFund. Temasek had co-led a US$90 million investment round in 2022 and also participated in a US$200 million round in 2023. After those rounds, reports indicated that Temasek held about 5 per cent of eFishery.

The destruction was not limited to cap tables. The scandal also hit the credibility of a wider ecosystem that has spent the past decade selling growth, disruption and inclusion to global capital.

The numbers that should have set off alarms earlier

The prosecution’s case, as outlined in earlier proceedings and reported by e27 on April 20, traced the alleged manipulation back to 2017, when eFishery’s cash balance reportedly fell to just US$8,142.

From there, prosecutors alleged that revenue manipulation continued from 2018 to 2024 as the company fought to sustain operations and maintain fundraising momentum. That timeline is the real nightmare here. This was not an isolated accounting slip or a bad quarter dressed up to buy time. The allegation was of a long-running distortion that coexisted with aggressive capital raising and a unicorn valuation.

State prosecutors had said Huzaifah and two other executives caused losses of more than US$4.1 million to the startup itself, while the broader investor damage reached around US$300 million. Two former executives, Angga Hadrian Raditya and Andri Yadi, also faced prison demands during the earlier phase of the case.

Also Read: How eFishery lost control of its narrative

That discrepancy between internal losses and investor wipeout matters. It shows how startup fraud does not need to drain every dollar directly to cause a catastrophe. Inflate performance metrics long enough, raise against the fiction, and the eventual collapse does the rest.

A founder’s defence, and a judge’s answer

During an earlier plea, Huzaifah argued the matter should not be treated as a criminal case. In remarks cited by Bloomberg, he said: “If, in leading a company scaling and evolving so rapidly, I am accused of making administrative errors, I am ready to be held accountable civilly.”

That defence now looks as revealing as it is unsuccessful. It reflects a mindset that has long existed, quietly, in parts of the startup world: that aggressive accounting is a by-product of speed, that investor expectations justify distortion, and that the boundary between operational chaos and fraud is somehow negotiable.

It is not.

In Huzaifah’s own earlier comments, the cultural rot was even more stark. “I knew it was wrong. But when everyone else is doing it and they’re still fine and never get caught, you start to question whether it’s really wrong.”

That sentence may end up being the most important line in the entire saga. It speaks to a broader ecosystem problem, not just an individual fall. Startup fraud rarely begins with a cinematic act of villainy. More often, it begins with pressure, storytelling, weak controls, investor enthusiasm, and a widening internal belief that numbers can be “adjusted” until the business catches up. Sometimes it never does.

The damage to Southeast Asia’s venture machine

The eFishery case lands at an awkward moment for the region’s startup market. Capital is already tighter than it was during the zero-interest-rate boom. Investors are demanding clearer paths to profitability. Public market exits remain uneven. Against that backdrop, a scandal of this size hardens every existing doubt.

The immediate effect is reputational. Limited partners and institutional investors will ask harder questions about how a company backed by blue-chip names could allegedly misstate performance for years without being stopped sooner. The secondary effect is practical. Expect tougher diligence, more forensic audits, milestone-based disbursements, stricter board oversight and far less tolerance for founder opacity dressed up as vision.

Indonesia, in particular, may feel the impact sharply. It remains Southeast Asia’s largest digital economy and still offers enormous startup potential. But eFishery has shown that size and promise do not immunise a market against governance failure. If anything, a large opportunity can sometimes make investors more willing to suspend disbelief.

That does not mean capital will flee Indonesia. It does mean the terms of trust are changing.

Not a Wirecard, but close enough to sting

Southeast Asia has seen governance failures before, but few with the symbolic weight of this one. Globally, comparisons to Wirecard and Luckin Coffee are unavoidable, even if the scale differs. In each case, inflated numbers met investor appetite, and narrative delayed scrutiny until reality became impossible to hide.

eFishery now joins that cautionary archive.

The difference is that Southeast Asia’s startup ecosystem is younger and still proving itself to global capital. That makes each scandal heavier. A blow in Silicon Valley can be framed as an outlier in a mature market. A blow in Indonesia risks being unfairly read as proof of systemic weakness, whether or not that conclusion is justified.

That is why this sentence matters beyond one founder’s fate. It tells the market that criminal accountability is possible. It also tells founders that “administrative errors” is no longer a phrase likely to rescue them once manipulated accounts and investor losses pile up.

The end of a fantasy

Visibly shaken, Huzaifah reportedly cried in court and embraced family members after the verdict, Bloomberg said. Humanly, it is a tragic image. Commercially and institutionally, it is also the end of a fantasy that too many in tech still cling to: that growth can excuse anything, and that governance can always be fixed later.

Also Read: eFishery founder held by Indonesian police over alleged embezzlement

It cannot.

eFishery was once sold as a symbol of what Indonesian innovation could become. It is now a warning about what happens when ambition outruns accountability. For Southeast Asia’s startup industry, that warning could prove more valuable than another unicorn headline. Painful, yes. Overdue, absolutely.

The post eFishery founder gets 9-year jail term, closing the book on one of SEA’s worst startup collapses appeared first on e27.

Across industries and geographies, technology has become embedded in nearly every professional’s workflow. A typical workday now starts and ends with notifications, as messages, meetings, dashboards, and updates compete for attention across multiple platforms. Yet as digital touchpoints multiply, many organisations are discovering that constant connectivity does not automatically translate into better collaboration, stronger alignment, or equitable participation.

According to recent research in APAC, while 69 per cent of organisations have adopted hybrid work models, six in ten employees are reporting moderate to high levels of burnout. Virtual and hybrid meetings, in particular, have emerged as the primary environment where employees report mentally disengaging while working.

However, the challenge is not simply about productivity. Increasingly, organisations are recognising that how digital collaboration systems are designed can shape who is heard, who participates, and who progresses in the workplace. The core challenge leaders face today is not access to tools, but understanding how to leverage them to foster genuine collaboration and engagement at scale.

A tech-integrated hybrid work environment offers endless opportunities for innovation, yet many teams only use digital tools to replicate their traditional, hierarchical ways of working. To effectively harness the power of both technology and human creativity, organisations must design their digital ecosystems intentionally to nurture a culture of collaboration that balances efficiency with meaningful human connection and inclusive participation.

When digital tools create friction instead of connection

In most organisations, employees juggle multiple platforms for communication, task tracking, and knowledge sharing, often with overlapping purposes and unclear ownership. Research shows that desk workers now use an average of 11 applications, up from 6 back in 2019. This complexity can lead to confusion and inefficiency, especially when collaboration norms are undefined, such as expectations around response times, meeting purpose and decision ownership.

But beyond inefficiency, poorly designed collaboration systems can also reinforce structural inequities in the workplace. Employees who are newer to organisations, working remotely, or located outside headquarters often have fewer informal opportunities to build visibility or influence decisions.

Also Read: How to launch collaborations that grow communities: A guide for Web3 founders

Without shared norms for collaboration, employees may experience constant interruptions and meeting overload, fragmented attention and reduced capacity for deep work, and a growing sense of isolation despite frequent digital interaction. Over time, unsolved collaboration friction can erode engagement and increase attrition, even in organisations that are otherwise digitally mature.

Why workplace culture must be treated as infrastructure

Collaboration does not fail because employees are unwilling to work together, but because leaders rarely design an intentional, consistent strategy for collaboration. In many technology-driven organisations, culture is still treated as a set of perks or values statements, rather than a form of organisational infrastructure that shapes how decisions, opportunities, and recognition flow.

An intentional collaboration culture means leaders actively shape how people work together by setting clear expectations and modelling consistent ways of working, instead of just providing tools. In practice, this type of culture gives employees clarity about: how decisions are made and communicated; where knowledge lives and how it is accessed and shared; and when collaboration adds value versus when focused, independent work should be protected.

When these systems are intentionally designed, collaboration becomes more inclusive and transparent. Employees have clearer pathways to contribute ideas, access information, and participate in decisions regardless of seniority, location, or background. This can spark motivation, as employees see the impact of their contributions towards shared goals, and can likewise strengthen retention, as research shows that employees with a higher sense of purpose at work are less likely to disengage or leave.

Rebuilding engagement and alignment with interactive digital tools

Unlike passive communication channels, which can distract employees from meaningful work, digital tools that use interactive formats prompt employees to engage and co-create. When embedded within a clear collaboration culture, these technologies can help restore energy, alignment, and participation across teams.

For example, tools that incorporate live polls, quizzes, and real-time feedback allow teams to align quickly on priorities and decisions while surfacing diverse perspectives across locations. Additionally, game-based elements can also reinforce learning through active recall of knowledge and skill practice among teams.

Also Read: Weathering the tariff turbulence: How AI and collaboration can lift SEA SMEs

Gamified interactions also create low-pressure opportunities for connection, making collaboration feel more human and inclusive. In hybrid environments, this approach helps level the playing field by ensuring voices are heard regardless of location, role, or visibility. 

Collaboration by design for a more equitable digital workplace

In an always-on, digitally mediated workplace, how people collaborate now shapes not only productivity, but also who has access to opportunity and influence within organisations. This is why collaboration by design is increasingly essential to building an engaged and equitable workforce. Leaders who intentionally design collaboration through clear norms, inclusive behaviours, and engaging digital experiences create environments where teams can perform sustainably without burning out.

The next evolution of collaboration is not about adding more platforms or flashy features, but about using technology thoughtfully to help people connect, contribute, and grow together with purpose. When workplace culture is treated as infrastructure, organisations are better positioned to build digital economies where participation and opportunity are more evenly distributed.

—

The post When collaboration systems break down in tech-driven workplaces and how to fix them appeared first on e27.

Across Asia Pacific, AI is already embedded in fraud prevention and anti-money laundering (AML). Most organisations use it daily and trust it; yet, many still cannot connect the outputs into a single, coherent view of risk.

For Tamas Kadar, co-founder and CEO of SEON, an AI command centre for fraud prevention and AML compliance, that gap defines the current phase of the industry.

“AI is already well established,” he says. “The issue is what happens when those outputs need to support decisions across the full lifecycle.”

Also Read: Building an anti-scam ecosystem is the key to a safer digital future

In many companies, onboarding, transaction monitoring, screening, and investigations still sit in separate systems. AI improves individual workflows, but without integration, it does not improve decision-making overall. The result? Fragmented visibility.

The real question is no longer whether companies use AI, but whether they can turn it into decisions that are consistent, explainable and fast enough for real-time environments.

Integration is becoming a growth advantage

The divide shows up most clearly in growth.

Higher-growth companies are far more likely to have integrated systems. Kadar stops short of claiming direct causation, but the pattern is consistent: companies that scale quickly tend to treat integration as core infrastructure.

“Fraud and AML complexity rises quickly as a business grows,” he says. “The companies that scale best integrate early, so complexity does not become a drag.”

Disconnected systems create friction. Teams spend time reconciling data, decisions lack context, and accountability becomes unclear. Integration reduces that friction and allows businesses to expand without letting fraud losses or compliance bottlenecks spiral.

What “unified” actually means

“Unified” is often used loosely. In practice, it means building a shared backbone.

Fraud and AML teams need access to the same customer context, decision logic, and audit trail. Risk signals, from behaviour to transactions, must feed into one system so AI can understand relationships between them, not just make isolated judgements.

This is difficult to implement. System complexity, talent shortages and incompatible data remain major barriers.

In Southeast Asia, the challenge is amplified by market variation. Payment systems, regulations and fraud patterns differ widely. A workable model is not a single rigid system, but a consistent core with local flexibility layered on top.

The biggest mistake: automating too early

Many companies move fast on AI but skip a critical step.

“The biggest mistake is automating before defining decision ownership,” Kadar says.

Also Read: Why AML compliance is becoming proptech’s biggest opportunity in 2026

Buying multiple tools or automating weak processes are symptoms of the same issue: unclear decision-making structures. Without clarity on how decisions are made and reviewed, automation simply accelerates confusion.

This becomes more serious as companies expand. Fraud and AML decisions need to be explainable, especially when they affect customers and compliance obligations across multiple markets.

AI does not remove that responsibility. It makes it more urgent.

When speed turns into operational debt

Startups often prioritise speed and patch systems later. In fraud and AML, that approach can break down quickly.

Operational debt becomes dangerous when temporary fixes start influencing high-stakes decisions: customer access, financial exposure or regulatory compliance.

The warning signs are straightforward: teams jumping between dashboards, different departments working from conflicting data, and leadership lacking a clear view of risk. At that point, the system is no longer supporting growth. It is slowing it.

There is also a timing problem. Fraud evolves quickly, but many systems are slow to deploy or adapt. Delays increase both costs and exposure to risk.

The challenge is not choosing between speed and structure. It is building systems that can do both.

AI is changing work, not replacing it

Despite expectations, AI has not significantly reduced headcount in fraud and AML. Instead, it has changed the nature of work. Detection has improved, but the overall workload has increased. More users, more transactions and greater regulatory scrutiny have expanded the scope of operations.

AI acts as a force multiplier. It supports analysis and decision-making, but humans remain essential for oversight, interpretation and accountability.

Most organisations still favour human-in-the-loop models. AI assists, but final judgement stays with people.

Accountability cannot be outsourced to AI

As AI becomes more involved in decision-making, responsibility becomes harder to define.

Kadar is clear: accountability does not sit with the model. It sits with the system around it. That includes data quality, decision rules, governance processes and leadership choices. When something goes wrong, the issue is not the algorithm alone, but the broader control environment.

Vendors must provide transparency. Teams must monitor outcomes. Leaders must ensure systems prioritise accountability, not just speed.

The uncomfortable truth

The industry’s biggest misconception is that AI fixes operational problems.“The uncomfortable truth is that AI exposes weak operations faster than it fixes them,” Kadar says.

Poor data, unclear ownership and disconnected systems become more visible when decisions accelerate. Without a solid foundation, AI simply amplifies existing issues. That is why AI adoption in fraud and AML is not just a technology decision. It is an operating one.

Also Read: Asia’s new cyber threat: AI that speaks your language

Companies that benefit most are not those with the most tools, but those with the strongest foundations: clean data, clear processes and governance that can scale.

Without that, AI does not create clarity. It creates faster confusion.

The post AI doesn’t fix broken risk systems; it exposes them: SEON’s Tamas Kadar appeared first on e27.