Energy & Technology

Taiwan has formally imposed export restrictions on Chinese tech giants Huawei Technologies Co and Semiconductor Manufacturing International Corp (SMIC), significantly escalating regulatory pressure on two companies central to China’s ambitions in artificial intelligence chip development. The restrictions were outlined in an updated version of Taiwan’s strategic high-tech commodities entity list, published quietly on the website of the International Trade Administration. The list now includes Huawei, SMIC and a number of their subsidiaries, marking a notable policy shift by Taipei. Until now, Taiwanese authorities had refrained from targeting specific Chinese firms despite existing limitations on certain semiconductor technologies. Under the current regulatory framework, Taiwanese companies must obtain government approval before exporting any goods or technologies to entities listed. This move is expected to partially sever Huawei and SMIC’s access to critical materials, construction technologies and equipment required for building advanced AI semiconductor fabrication plants — similar to those produced by Taiwan Semiconductor Manufacturing Co (TSMC) for global leaders such as Nvidia Corp. In addition to the parent companies, several of Huawei’s overseas units located in Japan, Russia and Germany were also included in the latest update. Neither Huawei nor SMIC provided comment in response to media queries made outside regular business hours. The decision comes against the backdrop of growing geopolitical tensions and trade restrictions that have reshaped the global semiconductor landscape. Huawei and SMIC, both already subject to export limitations by the United States, have been at the centre of China’s push to reduce reliance on foreign chip technology. In 2023, Bloomberg News revealed that a number of Taiwanese firms were discreetly involved in supporting Huawei’s efforts to construct a network of chip plants in southern China. The latest Taiwanese restrictions are expected to hinder those developments. TSMC, which supplies key clients such as Apple Inc and Nvidia, halted shipments to Huawei in 2020 following US-imposed export restrictions. Despite facing numerous curbs, Huawei and SMIC attracted international attention last year when they unveiled a domestically produced seven-nanometre chip, a move that surprised many in Washington. The restrictions coincide with rising cross-strait tensions. Earlier this year, newly inaugurated Taiwanese President Lai Ching-te referred to China as a “foreign hostile force” and introduced new countermeasures aimed at curbing external interference. Beijing, which claims Taiwan as part of its territory, has reiterated its intent to unify with the self-governing island — by force if necessary. -Bloomberg

TOKYO: Japan’s largest power generation company, JERA Co., has announced it will significantly increase imports of liquefied natural gas (LNG) from the United States through a series of new long-term supply agreements, marking a key milestone in Japan’s efforts to enhance energy security amid evolving trade negotiations with Washington. The Tokyo-based utility confirmed it has signed 20-year contracts with four LNG projects based in Texas and Louisiana. The agreements are expected to enable JERA to procure an additional 5.5 million tonnes of LNG annually from the US, almost doubling its current import volume of 3.5 to 4 million tonnes. “These new arrangements further our long-term strategy to build a diversified and resilient LNG procurement portfolio, supporting the stable and secure supply of energy for Japan and broader Asia,” the company said in a statement. The development aligns with Japan’s ongoing discussions with the United States over trade and investment cooperation. Fuel and agricultural imports have been key bargaining chips as Japanese officials seek to ease punitive tariffs introduced under US President Donald Trump’s administration. The US Department of the Interior welcomed the LNG agreements, describing them as a landmark in President Trump’s drive to expand American energy exports. Interior Secretary Doug Burgum said the commitment would generate nearly a quarter of a trillion dollars in economic impact and support over 50,000 jobs in the US LNG sector. JERA, which supplies roughly 30% of Japan’s electricity, plays a critical role in the country’s energy landscape. Japan, one of the world’s most energy-dependent economies and the fifth-largest emitter of carbon dioxide globally, imported 65.9 million tonnes of LNG in 2024. While Australia remained Japan’s largest LNG supplier last year, accounting for 41.6% of imports, Malaysia provided 15.6%, Russia 9.3%, and the US 8.4%, according to figures from Japan’s Ministry of Economy, Trade and Industry. The strengthened LNG partnership with the US comes on the heels of Japanese Prime Minister Shigeru Ishiba’s February commitment to increase annual Japanese investment in the United States to US$1 trillion, signalling a broader shift in bilateral economic ties. -AFP

BANGKOK : Thai energy company B.Grimm Power Public Company Limited and regional data centre platform Digital Edge DC have announced a joint investment of approximately US$1 billion to develop a hyperscale data centre in Thailand, signalling robust confidence in the country’s digital infrastructure potential. The forthcoming facility, with a planned capacity of 100 megawatts (MW), will be strategically situated in Chon Buri province, approximately 100 kilometres east of Bangkok. Commercial operations are scheduled to commence in the fourth quarter of 2026, as confirmed by B.Grimm Power CEO Harald Link and Digital Edge CEO John Freeman at a joint press briefing held earlier today. The initiative underscores the accelerating demand across Southeast Asia for artificial intelligence (AI), cloud computing, and other digital services, positioning Thailand as a key regional player in the digital economy. Usa Nuetap, Head of Data Centre Development at B.Grimm Power, revealed the company is also evaluating an additional US$1.6 billion in future investments to expand its data centre footprint, potentially adding a further 200MW in capacity. Thailand has rapidly emerged as an attractive destination for hyperscale infrastructure development, drawing billions of dollars in commitments from global technology leaders including Amazon.com Inc., Alphabet Inc., ByteDance Ltd., and Alibaba Group Holding Ltd. The Thai government, under Prime Minister Paetongtarn Shinawatra, is actively encouraging foreign investment in digital infrastructure through a range of tax incentives and policy support, with the aim of establishing the country as a regional hub for AI and advanced digital services. “Thailand stands out as one of Asia’s most compelling markets for digital growth, with surging demand for AI and machine learning,” said Freeman. “Thailand’s promotion of renewable energy also makes it appealing for data centre investment,” he added, citing the alignment with Digital Edge’s sustainability objectives. Digital Edge, backed by private equity firm Stonepeak Partners, currently operates 24 data centres across six Asian markets including India, Indonesia, Japan, Malaysia, the Philippines, and South Korea. -Bloomberg

SINGAPORE: New Relic, the Intelligent Observability company, today released its inaugural AI Unwrapped: 2025 AI Impact Report, offering a front-row view into how developer choices are fundamentally transforming the AI ecosystem. Drawing from comprehensive aggregated and de-identified usage data from 85,000 active New Relic customers over a year, the report reveals that developers are overwhelmingly embracing the largest general-purpose models, led by OpenAI’s ChatGPT, which accounted for more than 86% of all LLM tokens processed by New Relic customers.   “AI is rapidly moving from innovation labs and pilot programs into the core of business operations,” said New Relic Chief Technical Strategist Nic Benders. “The data from our 2025 AI Impact Report shows that while ChatGPT is the undisputed dominant model, developers are also moving at the ‘speed of AI,’ and rapidly testing the waters with the latest models as soon as they come out. In tandem, we’re seeing robust growth of our AI monitoring solution. This underscores that as AI is ingrained in their businesses, our customers are realising they need to ensure model reliability, accuracy, compliance, and cost efficiency.”  Developers Rapidly Shift to the Latest ChatGPT models  Enterprises are closely monitoring and adopting the latest innovations from OpenAI. The data shows ChatGPT-4o has been dominating more recently, followed by ChatGPT-4o mini. However, adoption of ChatGPT from version-to-version is occurring seemingly overnight as developers pivot toward newer, better, faster, and cheaper models. New Relic users have been rapidly shifting from ChatGPT-3.5 Turbo to ChatGPT-4.1 mini since it was announced in April. This shows that developers value cutting-edge performance and features more than savings.  Developers Start to Experiment with Unique AI Models Across Apps   In a countervailing trend, the findings also highlight increased model diversification as developers explore open-source alternatives, specialised domain solutions, and task-specific models, although at a smaller scale. Meta’s Llama emerged as the model that saw the second largest amount of LLM tokens processed by New Relic customers. In fact, New Relic saw a 92% increase in the number of unique models used across AI apps in the first quarter of 2025.   AI Monitoring Adoption Grows Steadily    Organisations need a unified AI monitoring solution that is easy to set up, configure over time, and provides an intuitive experience for any user—from DevOps to executives. Since its launch last year, enterprises have been adopting New Relic AI Monitoring at a steady 30% growth in usage quarter-over-quarter in the previous 12 months, giving them a solution to ensure AI model reliability, accuracy, compliance, and cost efficiency.   Python Dominates, but Java is Growing Quickly  With the most momentum, support, and tooling, the data shows Python continues to dominate AI applications, with customer adoption growing nearly 45% since last quarter. In terms of both the scale of requests and customer adoption, Node.js followed Python. However, Java usage has grown rapidly at 34% since last quarter, signaling more production-grade, Java-based LLM applications are to come from large enterprises.   The AI Unwrapped: 2025 AI Impact Report is available today. Read the full report.

Meta Platforms Inc is reportedly preparing a multibillion-dollar investment in Scale AI, a leading data infrastructure firm at the heart of artificial intelligence (AI) development. The potential deal, which may exceed US$10 billion (RM42.35 billion), would mark one of the largest private funding rounds in history and solidify Scale’s position as a cornerstone of the AI ecosystem. Founded by 28-year-old Alexandr Wang, Scale AI has become a critical player in the AI supply chain, focusing on one of the three foundational elements of the technology—data—alongside computing power and talent. Wang, who recently testified before Congress, urged US lawmakers to support a national AI data reserve, ensure sufficient energy for data centres, and avoid a fragmented state-level regulatory landscape. His influence in both technological and policy circles is rapidly growing. Although less widely known than OpenAI’s Sam Altman, Wang shares a similar trajectory. Both founded their companies nearly a decade ago and foresaw the coming wave of AI innovation. Like OpenAI, which secured a landmark investment from Microsoft, Scale has emerged as a pivotal enabler of next-generation AI. The company, most recently valued at US$14 billion in 2024, had already attracted Meta’s backing during a prior funding round. Scale’s evolution reflects the broader transformation of the AI sector. Originally focused on labelling images for autonomous vehicles, the company now curates vast volumes of textual data that underpin large language models such as ChatGPT. These models require immense datasets to train algorithms capable of emulating human reasoning. While Scale has faced criticism over its labour practices—particularly its reliance on low-paid contractors in countries like Kenya and the Philippines—Wang has defended the firm’s compensation levels, stating in 2019 that workers were paid within the 60th to 70th percentile of local wages. More recently, a spokesperson noted that the US Department of Labor had closed an investigation into the company’s labour practices. As synthetic, AI-generated data becomes more prevalent, Scale has adapted by enlisting highly qualified contributors. The company now employs professionals with PhDs, MBAs, law degrees and other advanced qualifications to guide AI through reinforcement learning techniques. These experts construct complex problem sets that help improve model performance in domains such as medicine, law, and international tax compliance. By early 2025, 12% of Scale’s workforce involved in model refinement held PhDs in fields like molecular biology, while over 40% held postgraduate degrees in various disciplines. The strategic use of expert-led training is helping Scale address the growing demand for sophisticated, human-like AI capabilities, particularly in the wake of the Chinese firm DeepSeek’s emergence as a major competitor in early 2025. Revenue at Scale has soared alongside the industry’s expansion. The company generated approximately US$870 million in 2024 and is projecting to more than double that figure to US$2 billion this year, according to Bloomberg News. Scale has also strengthened its ties with the US government, particularly in the defence sector. Wang, known for his strong stance on US-China tech competition, has become increasingly engaged with lawmakers concerned about national security. Michael Kratsios, a former Scale executive, now advises President Donald Trump on AI policy, underscoring Scale’s influence in Washington. For Meta, a deeper partnership with Scale represents both a technological and strategic advantage, allowing it to compete more effectively with rivals like OpenAI and Google while also expanding its footprint in defence-related AI. For Wang, the investment signifies a symbolic milestone. He once recounted that his entrepreneurial aspirations were fuelled by The Social Network, the film chronicling Facebook’s early days—bringing the Meta connection full circle. -Bloomberg