News related to
watch

The California-based server maker said the equipment had first been sold through an authorised reseller before being deceptively acquired and targeted for illicit diversion. Its statement framed the case as a successful intervention rather than a failure of internal controls, stressing that it had worked with Taiwan authorities to stop the movement of high-value systems used in artificial intelligence data centres.

The case has drawn attention because Supermicro sits at the centre of the AI infrastructure boom, supplying servers that integrate advanced graphics processors and networking components used to train and run large AI models. Such systems have become strategically sensitive as Washington, Taipei and allied governments tighten oversight of technology flows to mainland China, Hong Kong and Macau.

Taiwan prosecutors have been investigating suspects accused of falsifying export documents for high-end AI servers containing Nvidia chips. The servers were allegedly routed through Japan as part of an attempt to move restricted technology onwards to China-linked buyers. The seizure of 50 systems, though small compared with global AI server demand, is significant because it shows authorities are now pursuing hardware diversions beyond the better-known routes through Southeast Asia.

Supermicro’s response is also shaped by a separate and larger US criminal case that has placed the company under scrutiny. In March, US prosecutors charged three people linked to Super Micro Computer, including co-founder Yih-Shyan “Wally” Liaw, in an alleged conspiracy to divert about $2.5bn worth of AI server technology to China in violation of export controls. Liaw and another defendant were arrested, while a third defendant was listed as a fugitive. Super Micro Computer itself was not named as a defendant.

That case alleged a sophisticated scheme involving shell companies, false shipping documents, repackaging and dummy servers designed to mislead compliance checks. Prosecutors said controlled AI server technology was moved from the United States to Taiwan before being redirected through intermediary locations to reach Chinese customers. The allegations intensified questions over how manufacturers, distributors and resellers monitor the final destination of advanced hardware once it leaves controlled channels.

Supermicro has denied institutional wrongdoing and has said the conduct alleged in the US case violated company policies. The latest Taiwan-linked statement attempts to reinforce that message by presenting the firm as a partner to investigators. It thanked law-enforcement and legal officials, while saying industry and government must work together to strengthen safeguards and improve supply-chain visibility.

The timing matters for investors. Supermicro has been one of the most closely watched hardware suppliers during the AI build-out, benefiting from surging demand for rack-scale servers used by cloud providers, enterprises and sovereign AI projects. That growth has made export-control compliance a material business risk. A single enforcement case can raise concerns about sales screening, reseller audits, customer identification and shipment tracking, even when the company is not accused of direct wrongdoing.

Nvidia’s position adds another layer to the issue. Its most advanced AI processors remain central to global data-centre expansion, and restrictions on sales to China have forced chipmakers and server companies to navigate a shifting regulatory environment. Nvidia has repeatedly said it follows export rules and does not support unlawfully exported products. Its chief executive, Jensen Huang, has also underlined that server vendors must maintain their own compliance standards when handling systems containing restricted chips.

The Taiwan action highlights a broader enforcement shift. Authorities are no longer focused only on chip shipments; fully assembled AI servers are now a priority because they can be more difficult to trace after integration. Servers may contain restricted processors, specialised networking, memory and cooling systems that together create the computing power needed for frontier AI workloads. Once such systems are broken into multiple shipments or passed through resellers, detecting diversion becomes more complex.

The unit sits under the president’s office and is evaluating AI start-ups globally, with long-serving executive Zhang Endong leading the effort. The move places investment activity closer to Sea’s top management at a time when the company is trying to convert its scale across Shopee, Garena and Monee into a stronger position in the next phase of consumer internet competition.

The new team is part of a wider internal reorganisation aimed at accelerating AI adoption across Sea’s businesses. Zhang is also involved in projects examining how the technology can be deployed inside the group, from product recommendations and seller tools to game development and financial services. A company spokesperson declined to comment on the investment unit.

The initiative follows a series of signals from chief executive Forrest Li that AI has become central to Sea’s long-term strategy. Li told investors last year that a trillion-dollar market capitalisation could be possible if the company doubled down on artificial intelligence and made the right calls. That ambition is now being tested against tougher operating conditions, heavier competition and investor scrutiny over margins.

Sea’s pivot comes as its core businesses remain sizeable but increasingly contested. Shopee continues to dominate e-commerce across South-east Asia, where price-led competition, live commerce, faster delivery expectations and merchant services have raised both growth potential and operating costs. Rivals including TikTok Shop, Lazada, Amazon and Rakuten are intensifying their own use of AI to personalise shopping, automate seller support and streamline logistics.

Artificial intelligence is already being built into Shopee’s customer and merchant systems. Its applications include recommendation engines, search improvements, advertising tools and seller assistance functions. The strategic logic is straightforward: more effective targeting can increase conversion, better seller tools can deepen merchant loyalty, and automated service functions can reduce costs across markets with different languages, payment habits and logistics infrastructure.

Sea has also expanded its partnership with Google to develop AI-powered tools across e-commerce and gaming. The companies are exploring an agentic shopping prototype for Shopee, designed to move beyond basic search and recommendation functions towards digital assistants capable of helping consumers complete tasks across the shopping journey. Garena is expected to use AI to improve game development productivity, an area where faster content cycles and lower development costs can carry significant commercial value.

Financial performance has given Sea room to invest, though not without pressure. First-quarter revenue rose 46.6 per cent year on year to US$7.1 billion, while gross profit increased 40.7 per cent to US$3.1 billion. Net income reached US$438.2 million, up 6.7 per cent, and adjusted EBITDA rose 9.3 per cent to US$1 billion. The figures showed continued momentum, but also reflected the cost of defending market share and expanding services.

Shopee delivered gross merchandise value of US$37.3 billion in the quarter, up 30.2 per cent from a year earlier, while gross orders climbed 29.3 per cent to 4 billion. Revenue from the e-commerce arm rose 45.1 per cent to US$5.1 billion, though adjusted EBITDA fell to US$223.2 million from US$264.4 million a year earlier, underlining the margin trade-off behind its expansion strategy.

Monee, Sea’s digital financial services unit, has become another pillar of growth. Revenue rose 57.8 per cent to US$1.2 billion, while consumer and small-business loans outstanding reached US$9.9 billion. Non-performing loans more than 90 days past due remained at 1.1 per cent, a key metric as the company expands credit services across emerging markets where risk controls are critical.

Garena, once Sea’s dominant profit engine, also showed stronger momentum. Bookings rose 20.1 per cent to US$931.4 million, while quarterly active users stood at 666.5 million. The unit remains anchored by Free Fire, but Sea’s AI push suggests management sees broader potential in development tools, content generation, player engagement and operational efficiency.

Sea’s AI Centre of Excellence in Singapore adds another layer to the strategy. The centre is intended to strengthen local AI capabilities, support enterprise adoption and develop talent in areas such as engineering, research and product development. Its launch aligns the company with Singapore’s broader ambition to become a regional AI hub and gives Sea a clearer institutional base for experimentation.

The investment team could give Sea access to external innovation at a time when start-ups are moving faster than large platforms in areas such as autonomous commerce, generative game design, fraud detection, credit scoring and AI-enabled logistics. The challenge will be distinguishing durable business models from inflated valuations in a crowded funding environment.

Investors will be watching whether Sea’s AI spending produces measurable gains in revenue growth, user engagement and margin improvement. The company has already shown that it can rebuild profitability after years of expansion-led losses. Its next test is whether AI can become more than an efficiency tool and help create the next major growth curve for one of South-east Asia’s most valuable technology groups.

A draft framework under consideration would expand the role of monitoring agencies, usually SEBI-registered credit rating firms, and give them a stronger channel to flag diversion, delays or non-cooperation in the deployment of funds raised through equity issues. The proposals cover initial public offerings, rights issues, preferential allotments and qualified institutional placements, widening the regulatory lens beyond conventional IPO scrutiny.

Central to the plan is a proposal to lower the threshold for mandatory monitoring of issue proceeds from ₹100 crore to ₹50 crore. That change would bring a larger number of mid-sized fundraisings under independent review, particularly at a time when companies are using public markets not only for expansion but also for debt repayment, acquisitions, working capital and general corporate purposes.

The proposed framework would also require monitoring agencies to report directly to stock exchanges, rather than depend mainly on issuers for the onward dissemination of their findings. This would mark a significant shift in the balance of responsibility, as exchanges would receive a clearer view of whether the stated objects of an issue match actual fund utilisation.

Companies that fail to cooperate with monitoring agencies could face penalties, including fines of ₹50,000 for each violation. The measure is intended to address a weakness in the current structure, where monitoring agencies review deployment but have limited power when issuers delay information, provide incomplete data or fail to explain deviations from stated plans.

SEBI’s initiative comes after a period of heavy fundraising through equity markets, followed by a more cautious phase shaped by global risk aversion, pressure on domestic stocks and foreign portfolio outflows. Large companies and new-age firms have kept IPO plans alive, but several issuers have deferred listings as valuations became harder to defend. The regulator’s draft reflects a view that investor confidence depends not only on disclosure before fundraising but also on verifiable use of proceeds after capital is raised.

Under existing rules, companies raising equity funds must disclose the objects of the issue in offer documents and provide periodic updates on utilisation. Monitoring agencies submit reports on whether the proceeds are being used as promised. These reports are placed before audit committees and submitted to stock exchanges, but the enforcement burden largely remains with the issuer and the regulator.

The proposed changes would make monitoring more active and less dependent on company-led disclosure. Rating firms would be expected to examine bank statements, payment trails, board approvals and project progress more closely. They may also need to identify whether funds have been parked for long periods, shifted to purposes not approved by investors, or routed through related-party arrangements.

For investors, the changes could improve visibility over post-issue conduct, especially in cases where companies raise money for broad purposes and later alter business plans. Fund diversion has long been a concern in equity markets because public shareholders bear the risk when proceeds are used for objectives that differ materially from those advertised during the fundraising process.

For issuers, the proposals may raise compliance costs and increase documentation burdens. Smaller companies crossing the proposed ₹50 crore threshold would have to engage monitoring agencies and respond to periodic queries. Market participants expect some pushback from companies that argue existing audit committee and exchange disclosure requirements already provide sufficient oversight.

The regulatory argument is that tougher monitoring could reduce the gap between promise and performance. When a company raises money to build capacity, buy assets or reduce debt, the market prices the issue partly on those declared plans. If funds are later diverted or left idle without adequate explanation, minority shareholders have limited ability to intervene.

Credit rating agencies would gain a more prominent governance role under the proposed structure. Their task would move beyond checking utilisation statements to becoming a stronger early-warning mechanism for exchanges and investors. That may also bring questions over capacity, liability and conflict management, since the same agencies operate across ratings, surveillance and advisory-linked functions.

SEBI has been refining capital-raising rules across public issues, preferential allotments, rights issues and qualified institutional placements as it seeks to balance ease of fundraising with protection for minority investors. The latest proposal fits into a broader regulatory push to make market access faster while demanding greater transparency after money is raised.