Asia’s merchants are approaching a strategic inflection point: in agentic commerce, the contest is no longer won at the search-result page, but inside the machine’s decision loop.
Morph’s forecast may sound bold, yet the surrounding evidence from Salesforce, Adobe, and McKinsey suggests the foundations are already in place. Open protocols for identity, mandate, checkout and settlement mean AI agents can increasingly transact like autonomous buyers rather than glorified recommendation engines.
For Asian retailers, this is both liberating and brutal. It lowers historic frictions in cross-border trade, giving merchants in Jakarta, Bengaluru, or Ho Chi Minh City access to buyers with fewer checkout penalties. But it also weakens familiar moats. If agents optimise relentlessly for price, delivery, reviews and return terms, then brand premium, ad spend and marketplace ranking lose defensive power. Merchants built around search arbitrage and performance marketing may discover that the most valuable shelf space is no longer Google or Meta, but the agent’s preferred merchant graph.
The starkest line in the report is also the clearest strategic warning: carts will not be abandoned; merchants will be. In the agentic era, readiness becomes survival. Those who expose inventory, pricing and payments to agents early will capture the demand wave.
Regional
Grab pushes back on Indonesia exit reports amid GoTo merger talk: The super app firmly denied rumours of a market exit, reaffirming its decade-long presence and ~50% ride-hailing market share in Indonesia, as speculation swirls over a potential merger with local rival GoTo.
SG Enviro closes Series A to serve data centres, fabs in SEA: The Singapore-based water and wastewater treatment firm, backed by Emerald Ventures and SEEDS Capital, now targets the region’s fast-growing semiconductor, pharmaceutical, and data centre industries with modular solutions.
VinRobotics takes its humanoid robot to ICRA and COMPUTEX: Vingroup’s robotics arm debuted the VR-H3 at two of the world’s premier tech and robotics stages, positioning Vietnam as a serious contender in the global humanoid hardware race.
Cool Japan backs JumpStart’s Series C as AI vending machines scale: Japanese brand distributor Cool Japan Fund joins JumpStart’s Series C as the Indonesian startup’s AI-powered vending network surpasses 6,500 machines with 200% revenue growth, reshaping last-mile retail.
Chinese autonomous firm Neolix partners QuikBot in Singapore: The tie-up brings door-to-door autonomous delivery to Singapore, with Punggol Digital District serving as the initial testbed for Neolix’s self-driving vehicles operating under QuikBot’s logistics platform.
Indonesian crypto platform Floq raises US$11.3M in under a year: The platform, which amassed 1.8M users since launching, secured fresh capital to deepen its crypto offerings amid Indonesia’s growing retail digital asset adoption.
Singapore’s K25.ai secures US$6M commitment from NewGen: The AI startup closed the funding at an implied valuation of ~US$100M, with NewGenIVF backing its ambition to deploy AI-driven solutions across Southeast Asian markets.
Interviews & Features
Data, not hardware, is the real bottleneck in humanoids: Matrix Robotics CEO Allen Zhang says the scarcity of training data, not component costs, is slowing the industry. He targets 10,000 units by 2028 and predicts a shakeout that will leave only a few survivors.
‘We are confident enough’: Nicko Widjaja’s lawyers speak out: BRI Ventures CEO Nicko Widjaja faces an 11-year prosecution over a TaniHub investment decision. His legal defence team invokes the business judgment rule, arguing the case criminalises standard venture capital practice.
From a small town in Spain, Magnific quietly reshapes creative AI: CEO Joaquín Cuenca Abela built Magnific into a globally ranked creative platform, ranked 11th by a16z, while staying bootstrapped. India and Indonesia now form its largest user base outside the West.
International
Agentic commerce is quietly rewriting Asia’s e-commerce playbook: AI agents are emerging as autonomous buyers across Asia, with McKinsey, Salesforce, and Adobe data pointing to a US$3-5T opportunity that will fundamentally alter search, advertising, and brand discovery by 2028.
DeepSeek eyes US$7.4B in its first-ever fundraising round: The Chinese AI lab, valued between US$51.8B and US$59.2B, is drawing interest from Tencent and CATL as it seeks capital to scale its globally disruptive large language model business.
Uber commits ~US$500M to self-driving startup Nuro: The robotaxi partnership marks one of Uber’s largest autonomous vehicle bets, with Nuro set to expand driverless testing and move toward commercial passenger service under the Uber platform.
Crypto market faces liquidation spiral risk toward US$2T: Over US$370M in Bitcoin liquidations have already hit the market, with geopolitical triggers and leveraged positioning raising the risk of an extended downturn toward a US$2.17T total market cap support level.
A 65% probability explains Bitcoin’s next likely move: With US$789M in liquidations recorded and 11 consecutive days of ETF outflows, on-chain data and MicroStrategy’s first Bitcoin sale since 2022 point to a probable near-term directional move for the leading cryptocurrency.
SoftBank shares drop 10.6% amid broader tech sell-off: Shares fell sharply on renewed concerns about the conglomerate’s heavy AI investment exposure, despite the stock having gained ~70% year-to-date on bullish bets tied to its portfolio companies.
Korea Investment & Securities acquires 20% stake in Coinone: The move signals a major brokerage’s formal entry into digital assets, as South Korea’s financial incumbents accelerate integration with the country’s well-established crypto exchange ecosystem.
Cybersecurity
Japan’s top banks get access to Anthropic’s Claude Mythos: MUFG, Mizuho, and SMBC will deploy the model for cyber defence as part of a public-private working group, marking a significant step in Japan’s effort to embed AI into its national financial security infrastructure.
Singapore leads on security governance but struggles to enforce it: A JFrog report reveals that while Singapore ranks highly on security policy adoption, 54% of teams need over a week to produce compliance documentation and 18% lack any formal shadow AI policy.
TrendAI joins Anthropic’s Project Glasswing cybersecurity effort: The Singapore-based firm will use Claude Mythos Preview for AI-powered vulnerability detection, joining a coalition of 150+ organisations across 15+ countries working to raise global cybersecurity resilience.
Semiconductor
TSMC boss upbeat as AI boom shows no sign of easing: CEO C.C. Wei told investors that AI-driven chip demand remains exceptionally strong, prompting a raised revenue forecast and increased capital expenditure commitments for the world’s largest contract chipmaker.
LG Innotek to expand semiconductor substrate plant in Vietnam: The South Korean firm will begin construction of an expanded facility in Hai Phong in July, with completion targeted for May 2027, deepening the country’s role in the global semiconductor supply chain.
Cerebras says it works with every AI gear maker except Nvidia: The chip startup, whose CS-3 systems run 15x faster than GPU equivalents, counts AWS and OpenAI among its partners — positioning itself as the go-to alternative for AI compute buyers locked out of Nvidia supply.
AI
ChatGPT hits 1B monthly active users in record time: OpenAI’s flagship app reached the milestone faster than any consumer application in history, even as rival Anthropic’s Claude platform posts 640% year-on-year growth, underscoring a broadening of the generative AI market.
More US firms turn to China’s DeepSeek over pricey AI rivals: Ramp’s trending AI software list shows DeepSeek rising fast despite starting at just 0.3% corporate adoption in January 2025, as cost-conscious enterprises look beyond Anthropic and OpenAI, which still lead at 34.4% and 32.3% respectively.
Nvidia CEO Jensen Huang maps agentic AI’s new computing architecture: Huang outlined a distributed model where RTX Spark handles local inference and cloud handles scale, describing agentic and physical AI as the next frontier for Nvidia’s chip and software ecosystem.
Magnific bets on human-led AI infra for marketing and film work: The creative AI firm launched MCP, Flows, and Agents, enterprise tools designed to maintain brand consistency across campaigns, with a growing footprint across Southeast Asian markets in India and Indonesia.
Thought Leadership
The invisible shopper rewriting Asia’s e-commerce playbook: As AI agents begin making autonomous purchasing decisions, the US$500B GMV opportunity by 2028 will demand that brands rethink content, SEO, and advertising strategies built for human, not algorithmic, buyers.
The two human skills that make AI-native businesses work: In workplaces where AI handles execution, strategic thinking and systems design emerge as the irreplaceable human competencies — the ability to set direction and architect how work flows, not just complete tasks.
AI didn’t replace my team; it promoted them: Drawing on the contrast between Klarna and IKEA’s AI strategies, the author argues that teams augmented by AI don’t shrink; they move up the value chain, taking on more complex, judgment-heavy responsibilities.
How a 60-year-old cleaning supervisor built an AI agent in 6 weeks: The case challenges the assumption that AI is only for technical workers, showing that deep domain knowledge, not coding ability, is the true competitive moat for building useful, context-aware AI tools in any industry.
The AI trust gap: why SEA startups need proof before they scale: Southeast Asian enterprises remain cautious about AI adoption, demanding a credible “proof stack” of real outcomes before committing to scale, creating both a barrier and a differentiation opportunity for startups that can demonstrate tangible results.
I don’t know what I’m becoming as a marketer, and that’s the point: The author reframes uncertainty as a signal, arguing that marketers who shed the copywriter identity and become conductors of AI-powered workflows are best positioned to lead, especially in SEA’s leapfrog markets.
The always-on boardroom: when strategy stops being an event: AI enables continuous strategic recalibration, dismantling the traditional annual planning cycle. Boards that fail to adapt risk making decisions based on outdated assumptions in an environment where conditions shift in real time.
Great talent used to mean getting things done; now it means knowing what: The shift from executor to catalyst defines the new talent premium. As AI absorbs routine output, the most valuable employees are those who can identify what work matters, and why it does.
The talent question every founder needs to ask before they scale: Scaling prematurely with the wrong people is a compounding liability. The author argues founders must design for capability, not headcount, asking whether their team can adapt to the demands of the next stage, not just the current one.
When AI removes the work that taught us how to think: Delegating cognitive grunt work to AI risks hollowing out the developmental experiences that build judgment. The author warns that removing struggle from learning may produce workers who are fast but fundamentally shallow.
The talent reset: why AI is changing what makes people valuable: As AI commoditises knowledge retrieval, the new currency becomes judgment, discernment, and the ability to operate in ambiguity, shifting what organisations should look for and develop in their people.
AI is quietly redefining how much one person is expected to do: The productivity expectations placed on individuals are expanding as AI absorbs task volume, raising urgent questions about workload sustainability, burnout, and whether “doing more with less” is a strategy or a trap.
The credential trap: what I’ve stopped looking for in risk hires: Degrees and certifications no longer predict how someone handles uncertainty. The author makes the case for composite judgment, assessing how candidates reason under pressure, not how many credentials they can present.
AI didn’t just change the work; it changed who you should be hiring: Adaptability and learning agility now outrank domain expertise on hiring scorecards. The author argues that the best AI-era employees are those who can continuously reinvent their role as the tools beneath them evolve.
Why one person + AI is becoming a serious workforce model: AI crews, autonomous multi-agent pipelines managed by one individual, are shifting the workforce calculus. The author argues execution is now cheap; the scarce resource is the judgment to direct it well.
She spent 3 years as an old woman; your team spent 3 hours: Designer Patricia Moore’s radical empathy experiment becomes a blueprint for inclusive design in Asia’s ageing markets, where the silver economy represents a massive, systematically underserved user base that most product teams still fail to design for.
Greatness in the age of AI: why character outlasts competence: As technical skills get commoditised by AI, the author argues that integrity, resilience, and moral courage are becoming the true differentiators, qualities that cannot be automated, delegated, or replicated by any model.
Emotional intelligence makes AI training stick: Resistance to AI tools often masks identity threat, not skill gaps. The author introduces a six-step framework for helping teams emotionally process the shift, turning psychological safety into the foundation for effective and lasting AI adoption.
When everyone’s talking about OpenClaw and you’re not using it: The author cautions against reflexive adoption of trending AI tools, arguing that intentionality, knowing why a tool fits your workflow, matters more than keeping pace with the hype cycle in a market saturated with new releases.
The post Ecosystem Roundup: The rise of the machine consumer appeared first on e27.