For many companies expanding into Southeast Asia, hiring is no longer the hardest part. The bigger challenge is often what comes after: setting up a local entity, complying with employment regulations, running payroll and moving money across borders without unnecessary delays or excessive costs.

Singapore-headquartered TalentHero, co-founded by Oliver Lewis (CEO) and James Miles (COO), is looking to address these challenges through a new strategic joint venture in Thailand, alongside a proposed business partnership with Thai cross-border payments provider DeeMoney.

Also Read: 1982 Ventures joins Limited’s US$18.5M seed round to simplify cross-border business banking

The joint venture, established between TalentHero and Aswin Phlaphongphanich in his capacity as a private strategic investor, will focus on developing Employer of Record (EOR), recruitment, payroll administration and related workforce solutions for businesses operating in Thailand.

Separately, the new venture intends to explore payment partnerships, such as with DeeMoney, to support cross-border payments, international remittances and corporate payouts for its recruitment and EOR clients.

The agreement was signed in Bangkok on 2 October 2026, with Lewis and Miles representing TalentHero. Phlaphongphanich, who is also Co-Founder and CEO of DeeMoney, participated in his private investment capacity.

DeeMoney CFO Dylan Keota also attended, reflecting the parties’ intention to explore a commercial payments collaboration. The transaction was facilitated by Michael Gui, Partner at Copper Advisory.

At its core, the new venture targets a practical pain point. Companies want to hire in Thailand without necessarily establishing a Thai subsidiary from day one. At the same time, employers operating across multiple jurisdictions need reliable employment administration, payroll support and efficient payment processes.

“Building a team in a new market involves more than finding the right people. Employers also need reliable employment support and a practical way to manage payroll and payments,” said Lewis. “Through our new venture in Thailand, and by exploring commercial payment solutions with DeeMoney, we aim to make that journey simpler for businesses entering and growing in the market.”

Why EOR is becoming part of the expansion toolkit

Employer of Record, or EOR, has become a familiar model for startups and growth companies entering new markets. Under this arrangement, a business hires local employees through a third-party local employer, which handles employment administration, payroll and HR compliance. The client manages the employees’ day-to-day work, while the EOR handles the legal and administrative employment layer.

Also Read: TrueMoney and Fireblocks want Thai users on blockchain rails without them noticing

The model gained momentum during the remote-work boom, but its use in Southeast Asia is also tied to a more durable trend. Companies want to build distributed teams across multiple markets without carrying the cost and complexity of setting up local entities everywhere.

For Thailand, the proposition is timely. The country has long been a regional base for tourism, manufacturing, trade and services, but it is also positioning itself more strongly in digital industries. International companies entering the market need local hiring support, while Thai talent is increasingly part of regional and global teams.

TalentHero’s role in the joint venture will be to provide employment infrastructure and operational expertise. The company offers EOR, recruitment, global payroll and outsourcing services, combining talent acquisition with employment administration and operational support.

That matters because hiring is rarely a single transaction. A foreign company may first need to recruit a local country manager, then build a sales, operations or technical team, and later decide whether to incorporate locally. EOR can act as a bridge during that process, particularly for startups testing product-market fit or enterprises launching a regional function.

Where cross-border payments enter the picture

An additional opportunity being explored is the connection between employment services and international payment infrastructure.

DeeMoney, founded in 2017, specialises in international money transfers and cross-border payment services for individuals and businesses in Thailand.

Rather than participating as an equity shareholder in the new joint venture, DeeMoney is being considered as a separate commercial payments partner, with discussions focused on potential applications of its existing payment capabilities to the recruitment and EOR sectors.

These could include international remittances, cross-border business payments and payroll-related payout infrastructure, subject to applicable regulations, commercial arrangements and operational feasibility.

Thailand is both a destination and a source market for migrant workers, expatriates and regional business activity. For employees supporting families overseas, or companies managing payments to international contractors, suppliers or distributed teams, reliable payment infrastructure is an important operational consideration.

“Employment services and cross-border payments are closely connected, particularly for businesses operating internationally,” said Phlaphongphanich. “We see an opportunity to explore how DeeMoney’s established payment capabilities could complement TalentHero’s workforce solutions, helping businesses manage international payments more efficiently.”

The opportunity is not simply to offer more services under one roof. If commercially viable, closer coordination between employment administration, payroll processes and established international payment providers could reduce operational friction for companies building teams across Southeast Asia.

Any payment services arising from the proposed collaboration would remain subject to the relevant regulatory requirements and be delivered through appropriately licensed providers.

A crowded but growing market

The new venture will not be operating in an empty field. Globally, EOR and payroll platforms such as Deel, Remote and Oyster have built large businesses helping companies employ people across borders. In Asia, players including Multiplier and other HR-tech providers are competing to serve companies hiring across the region.

Also Read: Circle to acquire Tazapay for US$400M as USDC push moves into cross-border payments

On the payments side, DeeMoney operates in a similarly competitive landscape. Wise, Airwallex, Nium, Tranglo and traditional banks all serve parts of the cross-border payments market, while local fintechs and remittance providers compete on price, speed and coverage.

What could differentiate TalentHero’s new Thailand venture is the combination of international employment expertise, local market knowledge and potential access to established commercial payment infrastructure.

Global platforms often sell broad geographic coverage, but companies entering Thailand may still need local nuance. That includes employment administration, payroll timing, worker expectations, compliance processes and payment corridors that match actual business requirements.

Integration will nevertheless be important. EOR and recruitment are operationally demanding businesses, while international payments require appropriate licensing, risk controls and reliability. Any partnership must therefore balance customer convenience with regulatory compliance and operational effectiveness.

Thailand as a strategic growth market

For Southeast Asia’s startup ecosystem, TalentHero’s expansion reflects a broader shift in business infrastructure.

The first wave of digital tools helped companies sell online, communicate remotely and accept payments. The next wave is increasingly focused on essential operational functions: hiring, employment compliance, payroll administration and cross-border business payments.

Thailand offers an attractive environment for this approach. Its position within regional trade and investment flows, established industrial base and growing digital economy make it a relevant market for companies building international teams.

The joint venture also represents another step in TalentHero’s broader Southeast Asian expansion strategy, building on its existing regional operations and ambition to simplify international hiring and workforce management.

Also Read: Why investors often back Vietnamese startups more aggressively than Thai peers

For founders and operators, the appeal is straightforward. If they can enter a new market, hire efficiently, remain compliant and manage employment-related payments reliably, they can focus more resources on customers and growth.

The new Thailand venture is currently progressing through its establishment and operational preparations. The parties have not publicly disclosed its ownership structure or financial terms, and any commercial payment collaboration with DeeMoney remains subject to further discussions and applicable regulatory requirements.

As Southeast Asian businesses become increasingly regional, and international companies look deeper into the region for talent, employment infrastructure and cross-border payment services are becoming more interconnected.

Through its new Thailand venture, TalentHero aims to play a growing role in supporting that transition.

The post TalentHero, DeeMoney form Thai JV to connect hiring, payroll and payments appeared first on e27.

Every few weeks, another report tells the same story about Southeast Asian tech: capital has fled, deals have dried up, and the region’s startup engine has stalled. The most recent numbers make the case forcefully. Zero dedicated regional private equity funds reached final close in the first half of 2026.

Three pan-Asian mega-funds — EQT, Blackstone, and Bain Capital — absorbed US$39.2 billion, or roughly 85 per cent of all Asia Pacific PE capital raised in that period. Regional tech funding had already fallen 79 per cent between 2022 and 2024, from US$10.1 billion to US$2.2 billion, and exit values dropped a further 32 per cent in 2025.

Read those figures and the conclusion writes itself: the building has stopped. But that conclusion only holds if institutional capital is still the right instrument for measuring who is building.

The concentration problem sits inside the AI number too

Even the one bright spot in the data is thinner than it looks. Southeast Asia’s AI startups raised US$4.1 billion in 2026 through July, more than double the US$2 billion raised in all of 2025. That sounds like recovery, until you notice that Kling AI’s single US$2.8 billion Series D accounts for about 68 per cent of it. Strip that one round out, and native AI companies across the entire region raised roughly US$1.3 billion this year — a real number, but not the turnaround the headline implies.

So the capital that does exist is concentrating twice over: regionally, into a handful of pan-Asian vehicles, and within Southeast Asia’s own AI sector, into a handful of outsized rounds. Both trends are true. Neither tells you much about whether people are actually building.

Also Read: Light liquidations and flat funding: Is Bitcoin about to explode?

A second economy isn’t showing up on any cap table

While institutional funding was contracting, a different kind of activity was compounding. The market for AI-assisted, natural-language software building — often called vibe coding — is now estimated at US$4.7 billion and growing at roughly 38 per cent a year, with 63 per cent of active users being non-developers rather than trained engineers. The same reporting puts the typical span from a first idea to a paying customer at about four weeks — a cycle that requires no seed round, no pitch deck, and no line item in any PE report.

This is not a story about tools making bad builders look competent. It is a story about who gets to start. A product manager, a designer, or, increasingly, someone with no formal technical training at all can now ship and monetise a working product inside a month, entirely outside the instruments used to track “the startup ecosystem.” None of that activity registers as a data point in the reports currently used to declare the region’s tech sector shrinking.

I can speak to this from experience. I came up through the Normal Technical stream at school rather than a computer science degree, and spent years working as a paramedic before I started building software by prompting. What began as one product, a consumer signal-intelligence platform under construction at OnTheRice, eventually led to a board and product-architecture role at SCAVerse, a Singapore culture-tech venture.It does not counter the funding data. It is one example of a pattern those figures can’t capture: builders who exist before, and sometimes without, a term sheet.

The asymmetry now being closed isn’t only between founders

The same shift is playing out one level down, between institutions and the consumers those institutions serve. The AI consumer insights market is projected to grow from US$5.3 billion in 2026 to US$12.15 billion by 2034, with retail and e-commerce applications already accounting for about 45 per cent of that spend. What that money is buying, in practice, is the ability to point institutional-grade pattern recognition — the kind previously reserved for analyst desks and trading terminals — at ordinary consumer decisions: what a price move actually means, which claim in a press release holds up, which trend is a signal and which is noise.

Also Read: A funding round can improve your metrics without improving your company

Historically, that gap belonged to whoever could afford a terminal, an analyst, or a research subscription. It is the same gap that separated a funded founder from an unfunded one. Both are narrowing for the same reason: the cost of building serious analytical tooling has fallen faster than the institutions measuring “innovation” have updated their instruments.

Why this matters more than the funding headlines

If the only lens applied to Southeast Asia’s tech health is PE fund closes and disclosed equity rounds, the region will keep reading itself as being in retreat, even as real building activity — by non-traditional founders, using AI-native tools, serving consumers directly — may be at a multi-year high. That mismatch has consequences. Policymakers calibrate support programmes off these numbers. Investors calibrate conviction off them. A region that under-counts its own builders risks under-investing in exactly the layer where its next founders are already working.

The instruments didn’t just miss the recovery. They may be measuring an economy that is quietly becoming smaller than the one actually running underneath it — and the builders who never needed a term sheet to start are not going to wait for the definition to catch 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 Southeast Asia’s funding winter is real. Its building winter isn’t appeared first on e27.

Here is an uncomfortable finding for anyone building a 2027 marketing plan this month: audiences are warming to advertising, and marketers are growing less sure that theirs works.

That tension sits at the centre of Kantar’s Media Reactions study, published on 2 October and based on interviews with more than 800 senior marketers and 23,000 consumers. Only half of advertisers now feel they have the right balance between brand building and performance, down from 60 per cent a year ago. Confidence in the media mix slipped from 64 per cent to 57 per cent. And just 51 per cent say they are pulling actionable intelligence out of their data. Kantar’s own conclusion is blunt: the fix is sharper intelligence, not a bigger pile of data.

The hidden cost: The confidence gap

Call it the confidence gap: the distance between how much a team can measure and how sure it is about what to do next. It is easy to miss because every individual number looks healthy. Each platform reports its own engagement, its own attributed return and its own reach, and each one, read alone, says the money is working.

The trouble is that these numbers grade their own homework, and they often point the wrong way. Kantar’s creator research earlier this year found that platform engagement and real brand-building potential line up only about a third of the time. The number most teams use to pick creators misleads them two times out of three. Repeat that across search, social, retail media and video, and you get a dashboard that is busy, green and quietly uninformative.

Part of the gap is perception rather than performance. Kantar’s effectiveness data shows that channels working together drove 43 per cent of campaign performance between 2020 and 2025, up from 18 per cent before 2014. Marketers are getting better at making a fragmented mix pay. They just can’t see it clearly enough to defend it.

The moment it bites

Picture a regional head of marketing in Singapore in mid-October, three weeks from locking next year’s budget. Her quarterly review looks excellent. Paid social return on ad spend is up. Engagement on the creator programme has doubled. Cost per click fell in every market. Then the CFO asks the only question that matters: which of these moved revenue, and which can we cut?

Also Read: Why SEA’s retailers need to protect AI-powered customer experience with observability 

Nobody in the room can answer with confidence. Every channel can show a number that went up; none can show that the sales would not have happened anyway. So the budget rolls forward with a few percent shaved off everywhere, the same plan as last year, defended with the same slides.

That is the confidence gap turning into money. It is not one bad decision. It is a year of decisions nobody can explain.

The shift: A new channel nobody can measure yet

The gap is about to widen. Kantar finds that a net 75 per cent of marketers plan to increase investment in AI assistants in 2027, and 62 per cent expect generative AI to play a pivotal role in brand recommendations. Yet only 32 per cent of consumers use these assistants to research brands, and just 23 per cent trust what they recommend. Money is moving ahead of evidence.

The buy side already feels it. In the IAB’s outlook last month, 86 per cent of US buyers said they are changing how they measure media because of conversational AI and agents, and 45 per cent named comparing AI-driven and traditional customer journeys as their hardest measurement problem. Being found, cited and recommended in AI answers is turning into a goal in its own right, with no agreed way yet to count it.

What wins in 2027 is less measurement theatre and more signal discipline:

• Fewer, causal metrics. Pick the handful of numbers that track revenue or margin, and test them with holdouts instead of trusting each platform’s attribution.
• Prediction before spend. Score creative and audiences before the budget goes out, so the post-campaign report confirms a bet rather than explains it away.
• Category context. Read what competitors and audiences are doing across the market, not just your own funnel. A number only means something next to the market it sits in.

Why Southeast Asia feels it first

The confidence gap is wider here, for structural reasons. A regional team plans across half a dozen markets, several languages and a platform mix that changes at every border, from marketplaces to super-apps to messaging-led commerce. Global brands often hand local teams a direction from headquarters and a dashboard built for a single market, then ask them to defend spend line by line.

Also Read: AI chats are becoming the new starting point for online shopping

Agencies carry the same load with fewer people. An independent shop pitching against a network rarely loses on ideas; it loses on the depth of the read it brings into the room. That is why the insight stacks agencies build now matter as much as their creative. Whoever can explain the category, the audience and the competitive moves fastest earns the right to say what the numbers mean.

How the gap gets closed

The teams closing the gap are not adding dashboards. They are changing what they look at before they spend. As the founder of SOMIN, I see the same pattern across brands and agencies: read the category first, find the few signals that predict outcomes, and let those signals set the plan.

Two cases from our own work show how this plays out, and both are shared here as examples of the approach, not as proof the product works.

At Mothercare Singapore, the local team was tracking more than 4,000 category ads a year by hand and still guessing which content would land with Singaporean parents. Benchmarking the category by engagement per post rather than by volume showed that content about the real challenges of parenthood earned roughly 39 engagements per post, against about seven for festive gifting. With that read in hand, the team reported that time spent on research and ideation fell by 80 per cent, and that audience segmentation became three times more refined.

SAMY, an independent creative agency, used a similar approach to pitch a UK hospitality group against larger networks. Instead of leading with concepts, it walked in with a read of the client’s competitive set and an audience the category had left unserved, and won a multi-million-pound retainer. The agency now gets a click-through prediction on its creative before media money moves, which it reports at 78 per cent accuracy.

The three-number test

None of this requires a bigger martech stack. It requires the discipline to retire the metrics that flatter and keep the few that predict, and the confidence to explain that choice out loud: to the CFO, to clients and to the market. The leaders who do that in public, with a steady voice and the data behind it, will set the terms everyone else gets measured by.

So before the 2027 plan is locked: if you had to cut your dashboard to three numbers tomorrow, which three would you keep, and would your CFO pick the same ones?

—

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 Every dashboard is green. So why can’t anyone say what worked? appeared first on e27.

The cryptocurrency sector experienced a severe and sudden decline over the last 24 hours. Bitcoin fell 2.28 per cent to US$81,370.58. The broader digital asset ecosystem mirrored this downward trajectory, dropping 2.22 per cent to a total valuation of US$2.74 trillion.

An unexpectedly hot United States inflation report drove this widespread selloff across all major digital exchanges. The Bureau of Labor Statistics released the September Consumer Price Index data on October 14. This official government report showed the headline index rose 3.7 per cent year-over-year, beating the consensus forecast of 3.6 per cent.

Core inflation also surprised veteran investors by rising 0.3 per cent month-over-month, above the strict expectation of 0.2 per cent. These stubborn inflation numbers immediately altered investor sentiment and sparked a broad retreat from speculative risk assets. Retail traders and large institutional funds reduced their exposure to protect their capital.

Investors now anticipate a much more hawkish stance from the central bank due to persistent, sticky inflation readings. Higher interest rates directly reduce the appeal of non-yielding assets and force capital into safer government bonds. Policymakers will likely maintain elevated borrowing costs for a significantly longer period to combat the stubborn price increases across the real economy.

Traders eagerly await the upcoming Federal Reserve meeting scheduled for October 27 and October 28. The central bank committee will evaluate these fresh inflation metrics before making its final policy decision. Traditional equity exchanges and digital assets now move in near-perfect lockstep during major macroeconomic events. Bitcoin currently has a high correlation of 96.6 per cent with the S&P 500 index.

The total cryptocurrency space exhibits an even stronger 97 per cent correlation with the broader stock market. Professional traders increasingly view digital assets as high-beta extensions of traditional technology equities. This tight mathematical relationship means any negative macroeconomic news instantly pressures token prices and forces algorithmic selling.

Also Read: Could a hot inflation print send Bitcoin and Ethereum into freefall?

Technical indicators confirm the overwhelming bearish momentum across the entire digital asset landscape. Bitcoin recently broke below a very tight trading range and violated crucial historical support levels. Analysts previously identified the US$81,900-US$82,000 price range as a vital floor for the leading digital asset. The absolute failure to hold this specific area accelerated the downward price movement and triggered automated sell orders.

The 50-day moving average currently sits at US$82,636 and now acts as formidable overhead resistance for any attempted relief rallies. Aggressive short sellers completely control the immediate trend and actively push the price toward lower liquidity pools.

If the asset fails to reclaim the broken support zone in the coming days, aggressive sellers will likely drive the price down to test the US$78,000 mark. A deeper, more painful correction could easily push the valuation into the US$78,000 to US$79,250 support range, where long-term buyers might finally step in to accumulate more coins.

Excessive leverage in the complex derivatives market significantly amplified the recent drop in spot prices. The sudden macroeconomic shift forced the rapid liquidation of heavily leveraged trading positions across multiple global exchanges. Crypto derivatives platforms processed over US$71.16 million in Bitcoin liquidations within a single 24-hour period.

This massive forced selling represents a staggering 52.33 per cent increase from the previous trading session. Bullish traders absorbed the vast majority of this financial pain, as long liquidations specifically totaled US$59.15 million. These forced margin closures created a violent cascading effect that pushed the spot price even lower in a very short timeframe.

The derivatives sector currently shows a positive funding rate of +0.0048 per cent. This metric indicates that optimistic buyers still pay a premium to hold long positions despite the recent, severe price decline. Smart money traders need to see a complete stabilisation in open interest to confirm the market has successfully flushed out the excess leverage and established a firm base.

Also Read: Below US$82,620 Bitcoin could test US$81,154 next

The timing of this macroeconomic data release significantly worsened digital asset price action. The inflation report hit the trading desks during a period of notoriously thin weekend liquidity. Market makers and large institutional trading desks typically reduce their activity during Saturdays and Sundays. This lack of deep order book liquidity allows relatively small sell orders to cause massive price slippage. The combination of forced derivative liquidations and thin weekend order books created a perfect storm for volatility.

Retail traders watched their portfolios shrink rapidly as the market lacked the necessary depth to absorb aggressive selling pressure. This structural market vulnerability highlights the immense risks of holding highly leveraged positions during major macroeconomic announcements.

The total cryptocurrency capitalisation faces its own critical technical test on the daily charts. The aggregate valuation recently defended the crucial US$2.67 trillion support level against aggressive selling pressure from nervous participants looking to exit their risky bets.

Technical analysts recognise this specific price point as the exact 78.6 per cent Fibonacci retracement level from the recent major swing high. A sustained break below this crucial floor would expose the entire digital asset ecosystem to severe downside risks and panic selling. Aggressive bears would then target the yearly swing low situated precisely at US$2.6 trillion.

Conversely, a successful and decisive defense of this support zone could spark a powerful relief rally across the board. Confident buyers might then push the total market capitalisation back toward the 50 per cent Fibonacci level at US$2.77 trillion. The overall structural health of the digital asset ecosystem depends heavily on maintaining these aggregate valuation thresholds and demonstrating underlying buyer demand.

The immediate future of the cryptocurrency sector hinges entirely on the upcoming economic data releases. The Bureau of Labor Statistics will publish the highly anticipated Producer Price Index report later today, October 15, to provide further clarity on wholesale inflation trends.

Also Read: Bitcoin’s US$83,000 test: Can institutional demand hold the line?

This specific data point serves as the final major inflation metric before the central bank convenes for its crucial policy meeting. A softer reading could provide immediate relief to battered risk assets and spark a short-term technical rebound. Another hot inflation print will undoubtedly extend the current aggressive selloff and drive prices to new local lows.

I view the current market environment as a classic macro downdraft, in which stubborn inflation fears trigger a distinct flight from risk. Institutional buyers must step in to defend the US$81,000 support level and stabilise the bleeding across major trading pairs.

The trading action will soon reveal whether smart money views this dip as a prime buying opportunity or if more deleveraging awaits impatient retail traders. Traders must closely monitor daily closing prices to confirm the prevailing directional trend. A daily close below US$81,000 will confirm further downside toward lower support zones and validate the bearish thesis.

—

The post Institutional buyers face a massive test at the US$81,000 Bitcoin support level appeared first on e27.

At M360 ASEAN in Kuala Lumpur last week, the GSMA released its ASEAN Consumer Scam Report 2026. Nearly half of reported scam cases across six markets remain unresolved. Nearly nine in ten consumers said they had recognised AI being used in scams, from deepfakes to voice cloning. Julian Gorman, the GSMA’s head of Asia Pacific, summed it up: “Access alone is no longer enough.”

A month earlier, the GSMA’s Mobile Economy Asia Pacific 2026 named the rise of AI and declining digital trust as two of four forces shaping the region’s next phase. The share of ASEAN consumers who say they have been scammed rose from 31 per cent in 2024 to 45 per cent in 2025.

Across the region, operators are now moving AI into the customer app. The technology operators are racing to deploy is the same technology eroding the trust they need to deploy it. That dependency is missing from most business cases.

Trust is now a churn number

For years, trust sat in the compliance column. That is changing. The 2026 scam report found that people who have been scammed are more than twice as likely to switch accounts or providers as those who have not. The 2025 edition found that 36 per cent of consumers would switch financial services provider for stronger security built on telecom data.

Operators hold both sides of that equation. Their networks carry the calls and messages scammers rely on, and their data can help stop them. Some are already treating protection as a product: CelcomDigi sells a dedicated safety range alongside its mobile and fibre plans, and Globe has convened a Philippine anti-scam coalition with the Global Anti-Scam Alliance. Trust is already on the price list.

The risk sits with the operator

AI inside a telco app runs on the operator’s license and on a billing relationship that may be decades old. When a vendor’s model gets something wrong, the regulator calls the operator, and so does the customer.

Southeast Asia has mature rules for networks and, increasingly, for personal data. Consumer-facing AI inside an operator’s app is newer ground, and in practice much of its governance still lives in vendor contracts. That is a thin layer on which to rest a national operator’s reputation.

Also Read: Inside Singapore’s biggest telecom cyber defence operation

Subscribers have already written part of the spec

The most useful finding in the GSMA’s 2025 scam research was not a victim count. Consumer comfort with sharing data with conversational apps fell from 53 per cent to 40 per cent. In the same survey, 72 per cent said they were comfortable with limited, purpose-bound checks that use network signals to protect them.

Read together, those numbers describe the AI subscribers will accept. Not a chatbot that asks them questions. A system that works in the background, for a defined purpose, on as little personal data as possible.

Four tests operators can apply

In this context, trust is architecture, not assurance. Four design choices can be tested before anything goes live.

• First, the AI reads patterns, not people. There is no conversational surface between the subscriber and the model, so there is nothing to impersonate and nothing to phish.
• Second, the operator holds the identity key, and the intelligence layer sees only irreversible identifiers.
• Third, data is processed in-country.
• Fourth, the capability is exposed as an operator-controlled API: every call is logged, priced and revocable, and the operator can switch it off.

We had to build to these tests ourselves, because operator security teams kept asking the right question: if this sits inside our app, how do we know you will not walk off with our data? The answer that worked was not a promise. It was something their own teams could test before launch. That should be the standard for every AI vendor in this market. Do not ask for trust. Hand over something to test.

Also Read: How Agentic AI will create telecom’s first truly autonomous workforce by 2030

Breadth substitutes for intimacy

Personalisation is usually assumed to need more personal data. It does not have to. A system that sees wide behavioral signal across telecom, payments, content and gaming can personalise on patterns rather than identity. The more it sees of what people like you do, the less it needs to know about who you are.

Where trust is actually earned

Most subscribers do not decide whether to trust an operator by reading its privacy policy. They decide in small, frequent exchanges: whether points are credited when promised, whether a voucher works at the till, whether an offer feels useful or intrusive.

A user who watches food content and sees grocery vouchers surface first experiences AI as a service. The same signal used to push products they never asked for reads as surveillance. Every honored reward is a deposit. Every offer that feels like tracking is a withdrawal.

That exchange matters beyond churn. The GSMA estimates that 3.1 billion people live within mobile broadband coverage but do not use it. For many of them, the question is not whether the network reaches them. It is whether being online is safe enough to be worth it.

Operators planning AI launches for 2027 should be able to answer two questions before signing a vendor. Can our own security team test this before it goes live? Can we switch it off tomorrow? If the answer to either is no, the AI is not ready, however good the demo looks.

—

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 Telcos are racing to put AI in the app. Their subscribers have already said which kind they will trust appeared first on e27.