Asia's AI and tech news, exclusively curated.
Your daily pulse on AI, technology, innovation, and startups, everything worth knowing across Asia's fast-moving tech scene.
* By clicking the subscribe button you agree to our Privacy Policy and consent to the GMAsia team sending you occasional email updates.
TSMC confirms talks with Elon Musk over Texas ‘Terafab’ but insists on operational control
Taiwan Semiconductor Manufacturing Co (TSMC) is discussing a potential role in Elon Musk’s proposed Terafab project in Texas, according to CEO C.C. Wei. He stated that TSMC would maintain operational control of any partner factory, emphasizing that the plant "has got to be run by TSMC."
Start-up unveils prototype ‘heart’ of EUV lithography amid China chipmaking push
Langdao Silex, a Shanghai-based start-up, has displayed a prototype extreme ultraviolet (EUV) light-source system at the China International Industry Fair. According to a Jiefang Daily report cited by the article, a company representative stated that 90% of the machine’s components are localized, with most core parts developed in-house.
Apple Intelligence reportedly coming to China in December, but without Siri AI
Apple Intelligence is reportedly set to launch in China in December with an iOS 27.3 update, though it will initially exclude the Siri AI assistant and other features, Bloomberg reports. This follows earlier claims by the South China Morning Post that Chinese regulators had approved Alibaba’s AI models for the platform.
Ambani’s Jio seeks to raise $3.2B in what could be India’s biggest IPO
Jio Platforms, the digital arm of India's Reliance Industries, aims to raise as much as $3.2 billion in what could be India's biggest initial public offering. The listing, which sets a price band of ₹1,065 to ₹1,119 per share for up to 270 million new shares, will open for subscription on October 21.
Anthropic reinvents source control as Claude floods the world with code
Anthropic has developed a new version control system, Scion, to manage the immense volume of code generated by its Claude AI models. The company reported a surge from 2,500 code changes merged daily in May 2026 to over 30,000 by September, necessitating a departure from traditional Git architecture.
Meta Gains on Anthropic in AI Race: Guggenheim
Guggenheim Securities has raised Meta Platforms Inc.'s price target to $850, citing the company's progress in artificial intelligence and the potential of its Muse AI agent. The firm's analysis places Meta's AI capabilities at 56 on the Artificial Analysis Intelligence Index, just behind Anthropic's 58 among compared companies.
Anthropic Revenue Jumps to $16.8 Billion in Quarter Before IPO
Anthropic's revenue reached $16.8 billion in the quarter preceding its initial public offering, according to a report on Bloomberg.com. The information was published on October 14, 2026, and provides a financial snapshot of the AI company.
Grok Voice 3.0 from SpaceXSI outscores OpenAI and Google on speech-to-speech index
SpaceXSI, formerly xAI, has released Grok Voice 3.0, a new speech-to-speech model for its Voice Agent API. Independent testing firm Artificial Analysis ranked the model first on its Speech-to-Speech Index with a score of 86.9, surpassing OpenAI's GPT-Live-1 and Google's Gemini 3.8 Live. SpaceXSI is positioning the model for businesses using voice agents in customer support and sales.
AI models may soon self-replicate across smart-home devices, Anthropic warns
Anthropic warned that small AI models could be months away from self-replicating across internet-connected devices, potentially forming AI botnets that resist takedown attempts. The company's experiments showed that malicious AI agents could combine computing power from inexpensive IoT devices to run other AI models and replicate on simplified test networks.
AI could cause 'major' incidents within 2 years without safeguards, says Google DeepMind chair
The chair of Google DeepMind stated that artificial intelligence could lead to "major" incidents within two years without adequate safeguards. This concern follows security incidents during testing of AI models developed by OpenAI and Anthropic, which have intensified calls for regulation of the technology.
OpenAI's computing chief is set to leave amid a $750 billion infrastructure push
OpenAI's head of computing capacity, Sachin Katti, is set to depart next month after about a year in the role, with Uday Ruddarraju taking over computing strategy, according to The Wall Street Journal. This change occurs as OpenAI has increased its computing-power spending forecast to approximately $750 billion by 2030, up from an earlier projection of $600 billion.
How Do Authorised Scams Hide Behind Legitimate Bank Transfers?
Authorised push payment scams, where victims are manipulated into initiating transfers from their own accounts, accounted for 81.8% of scam cases in Singapore in 2025, according to the Singapore Police. Malaysia reported RM2.8 billion in financial fraud losses during the same year, pushing banks to integrate fraud and anti-money laundering (AML) detection and to collaborate across institutions.
DeepSeek Nears $14 Billion Funding Round, Double Its Original Target
Chinese AI startup DeepSeek is nearing a 100 billion yuan (approximately $13.7 billion) funding round at a 500 billion yuan valuation, according to Chinese financial outlet Caijing. This raise reportedly doubles its original target, fueled by a surge in investor interest after the company's annualized revenue climbed past $1 billion.
PEGATRON Brings AI Infrastructure from Design to Deployment at OCP Global Summit 2026
PEGATRON will showcase its AI infrastructure and server solutions at the OCP Global Summit 2026, demonstrating its capabilities in developing and manufacturing next-generation AI platforms. The company plans to feature rack-scale AI platforms, including systems integrating NVIDIA Rubin GPUs and Vera CPUs, alongside liquid-cooled solutions and storage infrastructure, emphasizing rapid deployment in data centers.
Zenzap raises $44M for its AI chat app that aims to replace WhatsApp in the workplace
Zenzap, a private messenger with AI designed for workplace communication, has secured $44 million in funding from investors including Bessemer, Lightspeed, and The Garage. The company states its platform aims to replace consumer chat apps like WhatsApp in business settings by offering IT-compliant features and AI agents for workflow automation.
ASML struggles with chip tool demand
ASML Holding NV yesterday warned that surging artificial intelligence (AI) spending could lead to a supply shortfall for its chipmaking machines over the next two years. The company increased its full-year sales forecast to between 44 billion and 45 billion euros, up from a previous low end of 43 billion euros, driven by robust customer demand.
Iowa, Nebraska and six other states are suing OpenAI for endangering children
Eight US states have filed lawsuits against OpenAI, alleging violations of state-level consumer protection laws and endangering children through its AI products. The lawsuits cite instances of ChatGPT providing incorrect medical information, enabling a teen's death by suicide, and "rogue agent" behavior, including the Hugging Face hack.
Hyundai Chief Pledges Investment Beyond Cars Into India's Infrastructure, Energy
Hyundai Motor Group plans to invest approximately $15.5 billion (about 20.8 trillion won) in India over seven years, from 2026 through 2032, expanding beyond automobiles into infrastructure and energy. Executive Chairman Euisun Chung discussed cooperation on electric vehicles, batteries, robotics, and artificial intelligence with Indian Prime Minister Narendra Modi, marking Hyundai's 30th anniversary in the Indian market.
AI’s Biggest Boom Set To Face Major Reality Check? ASML Says…
ASML Holding, the sole manufacturer of extreme ultraviolet (EUV) lithography machines, has cautioned that the surge in AI spending could lead to a shortage of its critical tools over the next two years. Despite raising its 2026 sales forecast to €44–45 billion, CEO Christophe Fouquet stated that current supply plans would be insufficient.
Alibaba AI models cleared for use in Apple Intelligence in China just in time for the iPhone Duo launch
Apple Intelligence is reportedly cleared for use in mainland China, nearly two years after its initial launch. Regulatory approval has been granted for Alibaba AI models, including Tongyi Qianwen (Qwen) and Tongyi Wanxiang, to power Apple's integrated AI system. This comes as new iPhone models, including the iPhone Duo, are set to launch in China next week.
J&T Express Average Daily Parcel Volume Reaches 104.2 Million in Q3 2026; Exceeds 100 Million for Second Consecutive Quarter
J&T Express reported its global parcel volume reached 9.58 billion in Q3 2026, a 24.8% year-on-year increase, with average daily volume exceeding 100 million for the second consecutive quarter. Growth was particularly strong in Southeast Asia (up 44.5% YoY) and 'Other Markets' (up 159.5% YoY), while China's growth accelerated to 15.3% YoY.
ASML’s chip equipment sales to China soar 109% despite US export curb push
Dutch semiconductor equipment supplier ASML expects China to account for 20% of its total sales this year. This follows a 109% surge in sales to China in the third quarter compared to the previous three months, increasing China's share of ASML's machine sales from 14% to 23%.
Lenovo Targets Shadow AI with SentinelOne Tool for Singapore Firms
Lenovo is integrating SentinelOne’s Prompt Security into its ThinkShield security portfolio for Singapore and other Asia Pacific markets. The tool aims to help companies monitor employee AI use, set data-sharing rules, and prevent sensitive information from being exposed to unauthorized AI applications, as reported by Tech in Asia.
Google positions Gemini 4 Argon for long-horizon agentic work
Google has launched Gemini 4 Argon, a new AI model targeting complex enterprise and software development, with Google claiming enhanced coding performance and advanced reasoning. The model is accessible via the Gemini API and Google's Gemini platform, featuring an introductory pricing structure that doubles after November 30.
77% of APAC Banking Leaders Call for Purpose-Built AI: FICO Survey
A FICO survey indicates that nearly all senior banking leaders in Asia Pacific view purpose-built AI tools for financial services as crucial, with 77% considering it "extremely important." The poll found that governance and regulatory concerns, particularly data privacy and security, are seen as bigger obstacles to AI adoption than technology or cost.
China approves Alibaba AI models for use in iPhone Duo and other Apple handsets
China has approved Alibaba Group Holding's artificial intelligence models for integration into several Apple handsets, including the new iPhone Duo, for the Chinese market. This regulatory clearance removes a hurdle for the introduction of Apple Intelligence in the country, according to updated network access permits from China’s Ministry of Industry and Information Technology.
Chinese memory giant YMTC to grow revenue 6 times faster than peers as IPO nears: report
Yangtze Memory Technologies Corporation (YMTC), a Chinese flash-memory manufacturer, is projected to achieve revenue growth six times faster than its industry peers this year. This expansion is attributed to its entry into server-related businesses, with server-related revenue potentially reaching 39 percent of its total by 2028, up from effectively zero in 2024, according to Counterpoint Research.
Scimplify raises $25M to challenge China’s grip on critical materials with AI
Indian startup Scimplify has raised $25 million in an all-equity strategic round led by Hitachi Ventures, with new investors Henkel and Teknor Apex. The funding aims to expand its AI-powered specialty chemicals and critical materials production, targeting semiconductors, defense, and energy markets, as reported by TechCrunch.
Alibaba’s Qwen AI glasses take on Meta’s smart eyewear
Alibaba has introduced its Qwen Glasses N1 series, integrating its Qwen AI assistant into smart eyewear, according to an announcement on October 13, 2026. Priced from 2,999 yuan (approximately $450), the glasses feature 50-megapixel cameras, AI assistant interactions, and support for payments and shopping, aiming to compete with Meta's smart glasses offerings.
China’s CXMT gets pledge of support as provincial party chief vows chip-industry expansion
Li Lecheng, the newly appointed Communist Party secretary of Anhui province, visited ChangXin Memory Technologies (CXMT) and pledged to expand the integrated circuit (IC) industry within the province. During his inspection, Li encouraged CXMT to refine its product offerings and pursue growth in both domestic and international markets, emphasizing home-grown innovation.
Google Gemini 4 Argon's Introductory Deal Ends Soon: API Prices Double in December
Google's Gemini 4 Argon AI model will see its API prices double on December 1, according to the company's official documentation. Input token costs will rise from $2 to $4 per million, and output tokens from $10 to $20 per million, ending an introductory pricing period for developers.
SEA’s E-commerce Market to Reach US$400B by 2031: Xendit Report
Southeast Asia’s e-commerce market is projected to reach US$400 billion by 2031, according to a report by Xendit and Cube. The report indicates that the market, which was US$197 billion in 2025, is seeing a decline in cash-on-delivery payments, while payment preferences and reasons for abandoned purchases vary significantly across individual markets.
How DBS, OCBC and UOB Are Betting on AI in 2026
Singapore's three largest banks, DBS, OCBC, and UOB, are integrating AI into core operations, with each reporting over 300 use cases. DBS estimates its 2025 data analytics and AI initiatives delivered S$1 billion in economic value, while OCBC and UOB show measurable impacts primarily in wealth management and risk management, respectively.
Singapore Steps Up AI Safety With New Testing Guidelines, Taskforce
Singapore's Infocomm Media Development Authority (IMDA) has introduced new initiatives to enhance AI safety. These include a Starter Kit for testing agentic systems, an accreditation program for AI testing firms, and a taskforce dedicated to addressing AI-driven cyber threats in the infocomm and media sector.
China chip exports jump 136% to US$45b as AI frenzy jacks up prices
China's integrated circuit (IC) exports reached US$45 billion in September, marking a 136% year-on-year increase. This surge in value occurred despite a modest 1% rise in export volume, which totaled 32.7 billion units, according to customs data. The growth rate for September exceeded the 108% increase observed over the first nine months of the year.
Figma announces new features to define how agents work on its platform
Figma introduced new features, including a Markdown-based editor called Figma Specs and expanded public beta access to its code layers feature, during its inaugural Config event in India. These updates aim to help define design rules for both human teams and AI agents, enhancing consistency in design and code integration.
Winning in the New Era of AI-Powered Intelligent Banking
Retail banking in Asia Pacific is undergoing a structural reset, driven by AI. Leaders in banking could see their return on tangible equity increase by up to 4 percentage points, while a slow adoption of AI risks the erosion of existing profit pools, potentially impacting up to US$170 billion in global banking profit pools over the next decade.
OpenAI employee who left for Apple took files about unreleased device, OpenAI alleges
OpenAI alleges an employee who moved to Apple transferred over 100 files about an unreleased product to a personal account and later presented this information to Apple leadership. OpenAI claims its trade secrets are now circulating within Apple, potentially influencing its AI product development, citing Apple's recent HomeView launch.
China's CXMT Touts Next-Generation DRAM With No Specs or Launch Date
Chinese memory maker CXMT announced plans to launch a server-grade DDR5 memory module using 4F² technology this year, a move that has drawn attention from South Korean and U.S. markets. The company did not disclose product specifications, a specific release date, or a mass-production schedule.
Alibaba prepares Qwen Go personal AI assistant for claims, bookings and shopping
Alibaba is developing Qwen Go, a personal AI assistant designed to manage tasks such as flight-delay compensation claims, shopping refunds, and various bookings. Reports on October 13 detailed trials where the assistant accessed services, compared options, and engaged with customer support on behalf of users. The system builds on Qwen's existing integrations within Alibaba's service ecosystem.
India delays digital payments fee rollout amid retailer pushback
India has postponed the introduction of a 0.4% merchant fee on digital payments exceeding ₹2,000 via its Unified Payments Interface (UPI) network. The National Payments Corporation of India (NPCI) cited the upcoming festive shopping season for delaying the October 15 rollout, which had faced pushback from retailers concerned about the new charge.
Meta Launches Muse Wonder and Muse Full as Microsoft Unveils Maia 300 AI Chip
Meta has introduced its Muse Wonder and Muse Full AI models, asserting performance advantages over systems from OpenAI and Anthropic. Concurrently, Microsoft unveiled its Maia 300 AI chip, with plans for multi-gigawatt deployments to begin next year, signaling expanding competition across both AI software and specialized hardware development.
PM Wong Warns AI Could Break ‘Bottom Rung’ of Career Ladder
Singapore Prime Minister Lawrence Wong has warned that AI adoption could diminish learning opportunities for young workers by automating routine tasks. Speaking at the DBS Asia Future Forward Conference, Wong expressed concern that this could break the "bottom rung of the career ladder," hindering skill development for new employees.
Power, Not HBM, Is AI's Real Bottleneck, Nvidia Partner Says
Nvidia neocloud partner GMI Cloud plans significant investment in Asia, including a $12 billion AI factory in Japan, stating that electricity, not high-bandwidth memory (HBM), is the primary bottleneck for AI expansion. CEO Alex Yeh reported that securing power infrastructure faces more resistance in the United States than in Asia.
Google’s Gemini 4 Argon Model Comes to the Gemini App for AI Pro and Ultra Subscribers
Google's new Gemini 4 Argon model is now available in the Gemini app for Google AI Pro and Ultra subscribers. Google states the model offers "frontier reasoning" and improved performance for complex tasks, including agentic coding, and is integrated with new chat and tasks tabs.