Singapore AI News & Daily Briefing

Chipmaker AMD says Malaysia will play a key role in Southeast Asia's push toward 'yotta-scale' AI infrastructure — massive computing power needed for next-gen AI. Enterprises are being told to rethink how they plan AI infrastructure as workloads shift toward continuous, autonomous AI processing.

Singapore's 11th OpenGov CXO Leadership Forum brought together senior leaders to debate how enterprises move from AI experiments to production-scale systems. Key insight: the bottleneck is rarely technology — it's organizational readiness, process redesign, and data quality. Speakers warned that layering AI onto broken workflows just automates inefficiency, and called for AI-native architectures rather than retrofitted legacy systems.

A Penang-based AI chip designer, SkyeChip Berhad, just listed on Malaysia's stock exchange in the country's biggest IPO in 16 years. The offering was oversubscribed 95 times, raising RM352 million. For decades, Malaysia handled semiconductor assembly and testing without owning the actual chip designs — this listing suggests that's changing fast.

A parallel forum session on data strategy revealed that most organizations have the technology but lack execution discipline. Speakers from Standard Chartered, Julius Baer, Liberty Mutual, and Denodo argued that enterprise data is fragmented across silos, dashboards often show yesterday's data instead of real-time insights, and AI initiatives fail because underlying data isn't AI-ready. The call: simplify legacy environments before adding more layers.

TSMC, the world's biggest chip manufacturer that makes processors for AI companies like Apple and Nvidia, is giving its employees more than a 30% raise in profit-sharing bonuses this year. CEO CC Wei made the promise at a staff town hall, responding to online grumbling about bonus sizes. The company posted NT$572.5 billion (S$23.3 billion) in Q1 earnings—more than double what it made two years ago—and its gross margin sits at a juicy 66%. This comes as Samsung's workers also won massive bonuses after threatening to strike, showing how chip industry workers are now demanding a bigger slice of the AI boom.

Two major data centers landed in Malaysia's pipeline within days of each other this month. The facilities are being designed differently now — built specifically to handle the crushing demands of AI workloads, not just traditional cloud storage.

China has started running an underwater AI data center near Shanghai, housing nearly 2,000 servers in pressure-resistant modules next to offshore wind turbines. The setup uses seawater for cooling, cutting energy costs. It's a bold bet on efficient, edge-located AI infrastructure away from crowded cities.

OpenAI Daybreak brings AI cybersecurity into code review, threat modelling, and patching. Tanium said Daybreak pressures Malaysia’s monthly patching cycles. OpenAI has announced Daybreak, a cybersecurity initiative that uses AI models and coding agents to support software defence for enterprise software development. The company launched Daybreak on May 11th, according to details cited from OpenAI’s […] The post OpenAI Daybreak and the patching cycle appeared first on TechWire Asia .

Google used its annual I/O conference to spread AI across its entire product lineup. New features hit Search, Gemini, and shopping, a new Gemini 3.5 version launched, and smart glasses got an update. Google also quietly lowered some AI subscription prices — good news if you've been paying for Gemini Advanced.

Huawei unveiled its Tau Scaling Law at a major circuits conference in Shanghai, claiming it can achieve chip density equivalent to 1.4nm by 2031—beating the limits of traditional Moore's Law scaling. The company's LogicFolding architecture will debut in Kirin chips later this year, backed by six years of mass production across 381 chips.

OpenAI launched Daybreak, an AI-powered cybersecurity system that helps developers automatically find and fix software vulnerabilities during code review and patching cycles. The system is designed to speed up what companies do manually, especially in regions like Malaysia where patching schedules lag behind global standards.

European security firm ESET announced a $40 million investment into AI-powered cybersecurity. The focus: protecting AI systems themselves from attacks and building new tools to process security data at scale. They're also flagging risks from OpenClaw vulnerabilities.

Huawei is working on a new approach to chip design that could make semiconductors more efficient. The plan, called Tau Scaling Law internally, focuses on reducing the distance data needs to travel inside processors—which could lower costs and boost performance. Investors in China are excited because it shows Huawei pushing forward despite US sanctions limiting its access to cutting-edge manufacturing tools.

Kong Inc., a US firm that builds API and AI connectivity tools, has launched its Kong Konnect platform in Singapore. Singapore organizations can now host the control plane locally and access dedicated cloud gateways, making it easier to safely connect AI models to their data and applications. The move reflects growing regional demand for enterprise tools that manage AI integrations at scale.

A company CEO announced significant job cuts blaming AI automation, only to receive threatening messages targeting his family. This reflects growing tensions as workers react to AI-driven layoffs. While companies see cost savings through automation, many employees feel disposable and unheard, leading to real-world hostility toward leadership.

Singapore organisations rank among Asia Pacific's most advanced on AI governance policy, but a new JFrog report exposes consistent failure to translate those policies into actual enforcement. This leaves enterprises exposed, particularly around software supply chain security. The gap between having AI policies and actually enforcing them is a real risk for businesses.

BDx, a Singapore-based data center developer, has locked in 1.2 gigawatts of power capacity in Indonesia through a deal with state utility PLN—the biggest power commitment ever secured by a data center operator in Indonesia. The agreement covers three facilities including the CGK4 AI Campus in West Java, which is Indonesia's first NVIDIA DGX-Ready-certified site built for AI training with H100-class GPUs. A second site in South Jakarta will support liquid-cooled, high-density AI compute including the next-generation NVIDIA GB200. For everyday users, this matters because AI services—from chatbots to cloud features—run on data centers that need enormous, reliable power. This deal signals that Southeast Asia is becoming a key battleground for AI infrastructure.

Vietnam's biggest conglomerate Vingroup just unveiled humanoid robots at two major global tech events — ICRA 2026 in Vienna and COMPUTEX Taipei. Two subsidiaries built separate models: VinRobotics showed the VR-H3, a third-generation industrial robot with remote operation via VR headset, while VinDynamics launched Dyno, a household and security assistant. Both companies built everything from scratch — mechanical parts, AI, power systems, and actuators — marking a rare Vietnamese tech debut on the world stage.

Google is rolling out AI-powered voice scam detection for Android phones before the end of June. The feature uses AI to analyze calls in real-time and alerts users when suspicious patterns are detected. This builds on existing scam call protections already available on Android devices.

Singapore's Urban Redevelopment Authority is showcasing practical AI tools for city planning at a new exhibition. One standout project uses sensors on seniors' bodies and robots to identify environmental hazards that could cause falls — helping planners design safer neighbourhoods for an ageing population. Other tools include a handheld scanner that detects building defects in 30 seconds using AI, and cameras that flag illegal dumping and unsafe parking in real time.

MedPlanner, a health tech company based in Kuala Lumpur, has launched two AI products: HEMI Health, an AI medical scribe with legal protections for doctors, and AskHEMI, a multilingual app helping patients manage chronic conditions. Both run on MedPlanner's proprietary health AI system.

Vietnam has become a key hub in Nvidia's sovereign AI strategy. Local firm FPT built a population-scale AI training dataset for Vietnam, while Viettel is developing a national legal AI app using Nvidia's models. Singapore is among the countries included in Nvidia's sovereign AI network, alongside South Korea, India, Japan, and others.

OpenAI, Anthropic, and xAI are all actively hiring in Singapore, with OpenAI leading with 22 open positions at its Singapore office. The city-state has become a key base for frontier AI labs including Google DeepMind and AMI Labs, reflecting Singapore's ambition to be Asia's AI hub with nearly US$780 million committed to AI research over four years.

Nvidia is collaborating with Intel and Taiwan's MediaTek to build new laptop chips that combine CPU, GPU, and AI processing on a single chip. The chips target AI PCs and will appear in devices from Dell and Lenovo. Nvidia is also investing $5 billion in Intel stock as part of this push.

Your colleagues might be secretly using AI tools at work that could be leaking company secrets. A new workplace trend called "Shadow AI" is emerging across Asia, where employees download AI chatbots and automation tools without IT approval. In Singapore, 68% of workers use AI frequently on the job, but only 14% stick to company-approved tools. This means sensitive data from your workplace emails, documents, or customer records could be feeding into public AI systems without your boss's knowledge. Companies are now turning to DNS monitoring and clear AI usage policies to catch these "shadow employees" before they accidentally expose private information.

Hyundai is rolling out Boston Dynamics Atlas humanoid robots across its car plants, targeting 30,000 robots. The company set up new units for software-defined factories and trade strategy amid US tariff pressures. Atlas handles material handling and parts sorting, with first real-world deployment at a Georgia plant from 2028.

A new study finds that the Asia-Pacific region leads the world in AI adoption but is struggling to find enough skilled workers. For everyday users, this means companies may rush to roll out AI tools, but quality and support could suffer without enough talent.

Singapore has climbed into OpenAI's top five markets globally for Codex usage — the AI coding tool now has 4 million weekly active users. OpenAI is working with big consulting firms like Accenture, Capgemini, and PwC to bring Codex to enterprise teams. It's a sign that AI coding assistants are gaining real traction among professionals here.

Marketing agencies are rethinking their approach as AI makes generic content cheap. Instead of acting as 'order-takers' who execute briefs blindly, the best partners now co-create strategy with clients from the start. Tech in Asia Studios describes its own model as an 'ecosystem interpreter' that embeds teams within client strategy rather than working from outside.

UN researchers warn that data centres powering AI will double their electricity and water use by 2030. Last year they consumed 448 terawatt-hours of power (more than Saudi Arabia) and 4.5 trillion litres of water. Water use alone is set to reach 9.3 trillion litres. The AI tools you use daily run on real physical infrastructure with significant environmental costs.

Researchers have created AI-powered computer worms that can spread between systems without human interaction. Unlike traditional viruses that need users to click something, these worms autonomously infect devices, copy data, and jam communications. Security experts are calling this a paradigm shift in cyber threats.

Microsoft has launched its own AI models to compete directly with OpenAI and Anthropic. The move marks a major shift as Microsoft — a key backer of OpenAI — develops independent AI capabilities.

Nvidia announced RTX Spark, a new processor for Windows laptops, at GTC Taipei during Computex. The company also detailed Vera and Vera Rubin chips designed for AI data centers, alongside infrastructure updates. RTX Spark brings AI capabilities directly to consumer Windows PCs.

Asus is using Nvidia's DSX platform to create 'digital twins' of data center setups—virtual simulations of power, cooling, networking, and storage—before building physical infrastructure. This helps companies deploy AI data centers faster and with fewer costly mistakes.

Russia's Sberbank is offering its AI model to developing nations, positioning it as a tool to close the digital gap. This is part of a broader geopolitical competition over AI influence—Western and Chinese AI dominate, and Russia is now pitching its technology as an alternative for countries seeking more options.

Enterprise companies are now being judged on how efficiently they run AI workloads as data center energy consumption surges. The focus is shifting beyond infrastructure to the logistics and systems that support AI computing, with pressure to scale without increasing carbon emissions.

Meta is rolling out AI-powered 'business agents' designed to handle routine enterprise tasks like customer service, scheduling, and internal communications. These tools aim to automate daily operations, potentially reducing workloads for small business teams.

Alibaba just dropped preview versions of its next Qwen AI models ahead of a big launch in Hangzhou. Early tests rank them among the world's top AI systems, beating other Chinese labs but still behind OpenAI, Google, and Anthropic. The company is burning billions in profit for market share — free cash flow dropped to a $2.5 billion outflow last quarter — but AI product revenue has grown at triple-digit rates for 11 straight quarters.

# Google unveils Gemini AI for multi-step Android tasks [News](https://www.techinasia.com/)Premium * [News](https://www.techinasia.com/) * [Jobs](https://www.techinasia.com/jobs/search) * [Companies](https://www.techinasia.com/companies) * [Events](https://www.techinasia.com/events/live) * [About](https://www.techinasia.com/about) * [advertise](https://www.techinasia.com/advertise?utm_source=techinasia.com&utm_medium=website&utm_campaign=intl-navbar) * [Asia Economic Summit](https://aes.techinasia.com/?utm_source=navbar-tia-homepage&utm_medium=website&utm_campaign=events-flagship-aes) * [Sign Up](https://www.techinasia.com/auth) * [Log In](https://www.techinasia.com/auth?ref=nav_login) [Subscribe](https://www.techinasia.com/subscription) [I'm fundraising](https://www.techinasia.com/compani

Major AI chatbots like ChatGPT and Microsoft Copilot are spreading pro-Russian propaganda as if it were verified fact, according to a NewsGuard investigation. The problem is worse in non-English languages. Chatbots tend to repeat whatever information is most widespread online, not what's most reliable. Researchers warn this could get worse as more people use AI to get their news.

Taiwan is reportedly exploring armed robot dogs for military defense against potential threats from China. The story taps into a broader global trend of nations deploying AI-powered autonomous systems for combat and surveillance roles.

UK publishers can now choose not to appear in Google's AI search summaries, thanks to a new rule from the Competition and Markets Authority. The regulator says this will give publishers more power to negotiate content deals with Google. Many websites have seen big drops in traffic since Google started placing AI summaries at the top of search results. Google controls over 90% of UK search and has nine months to implement the changes, which also require proper attribution when content appears in AI results.

Taiwan's Asus showcased its 'AI City' concept and new laptops powered by Nvidia chips at Computex, a major tech showcase. The AI City appears to integrate AI into urban environments, while the laptops target creative professionals and gamers needing heavy AI processing power.

The US FDA is evaluating an AI tool that predicts liver injury risk from medications before they reach patients. If approved, doctors could use it to flag potentially dangerous drug interactions early, improving drug safety testing.

Court documents reveal Microsoft was targeting a $92 billion return on its early $13 billion OpenAI investment, with the deal granting Microsoft a 27% stake. OpenAI paid Microsoft $493.8 million in revenue share in 2024, rising to $865.8 million by late 2025. The disclosures came during Elon Musk's lawsuit against OpenAI and Microsoft over its shift to a for-profit model.

SaaStr, a B2B software community, built custom AI tools to handle the boring sales tasks humans skip. The trick: start with ONE annoying workflow, connect it to your existing database like Salesforce, and let the tool grow as people actually use it. They booked 614 meetings worth $85K each by automating scheduling links for leads humans deemed not worth their time. The warning: AI can repeat mistakes thousands of times before anyone notices, so human oversight is non-negotiable.

Cohesity, an AI-powered data security company, has patented a new RAG (retrieval-augmented generation) architecture specifically designed for backup systems. This is a technical patent for enterprise data security platforms.

BDx Data Centers has locked in 1.2 gigawatts of electricity from Indonesia's state utility PLN—the largest power allocation held by any single data center operator in the country. That's enough juice to run roughly a million typical homes. The power will support three Indonesian sites used for AI training and cloud workloads, including facilities running Nvidia's H100 chips. The deal signals how seriously operators are competing for energy resources in Southeast Asia, where AI facilities are hungry for reliable electricity. BDx expects to consume all the capacity within two to three years as demand from Indonesian enterprises deploying GPUs grows.

Billionaire investor Ray Dalio breaks down what he sees as fundamentally different philosophies driving AI development in China versus the United States. China's approach tends toward state coordination, while the US leans on private market competition. Dalio suggests this ideological clash will be a major driver of geopolitical and economic outcomes in the decade ahead.

AI company Anthropic agreed to pay Elon Musk's xAI $1.25 billion per month through 2029 for computing power from its data center. This massive deal shows how AI firms are scrambling for hardware and could lead to faster, more reliable AI services for users.

United Airlines is building an AI system that will automatically explain flight delays to passengers in plain language—without any human involvement. The airline's 'Every Flight Has a Story' program already uses generative AI, but currently combines it with human editors. The goal is full automation: the AI will text customers about what's happening with their specific flight, including sending maintenance videos. United also uses AI for its ConnectionSaver tool, which decides whether to hold departing planes for connecting passengers—it has saved nearly 54,000 connections in Denver alone. CEO Scott Kirby calls this infrastructure that will help United run its entire operation more efficiently.

Elon Musk's SpaceX and Tesla are proposing a $55 billion semiconductor plant in Texas called Terafab, potentially expanding to $119 billion. The idea: build captive chip supply for AI and computing needs instead of relying on Samsung or TSMC. If it goes ahead as planned, it would rival TSMC's Arizona investment and even exceed the US CHIPS Act budget.

This is a general discussion piece exploring how AI is being used in sports broadcasts, gaming, music, and events. It covers things like AI-generated commentary, personalized highlights, and virtual reality experiences. Nothing particularly new or groundbreaking.

TSMC plans to build advanced chip packaging facilities in Arizona by 2029, easing bottlenecks in US AI chip production. The facilities will support CoWoS and 3D-IC packaging used in AI chips for Nvidia and others. TSMC has a deal with Amkor to provide packaging and test services for wafers made at its Phoenix plants. This marks a shift toward a more self-sufficient US chip supply chain.

Nvidia has poured US$2 billion into chipmaker Marvell Technology, deepening their partnership on custom AI semiconductors. Marvell makes chips for Amazon Web Services and other hyperscalers. The deal links Marvell's optical interconnect technology with Nvidia's NVLink Fusion platform, designed to handle faster data movement inside AI data centers. Jensen Huang reportedly said Marvell could reach a US$1 trillion valuation. Alphabet and Meta are expected to spend over US$630 billion on AI infrastructure this year.

Intel shares surged 58% over nine consecutive days, the chipmaker's best run since the 1970s. The rally was driven by partnerships with Google (Xeon 6 CPUs for AI), Nvidia (US$5B investment and custom chip development), and Elon Musk's Terafab project in Austin for SpaceX, xAI, and Tesla. Intel also bought back its Ireland chip plant for US$14.2 billion.

A US trade agency launched a patent investigation against memory chip makers SK hynix and Kioxia over 3D chip designs. If imports are banned, it could disrupt the supply of memory chips used in AI servers, potentially slowing down AI development and raising costs.

Nvidia is building a new campus called Constellation in Taipei, spanning nearly four hectares, with construction starting in the next couple of months. CEO Jensen Huang unveiled the plan at an employee meeting on May 28. The company already has an office in Kaohsiung and is hiring heavily in Taipei as its Taiwan supply chain grows. Huang said Nvidia sees growth in agentic AI and that physical AI will change manufacturing in Taiwan, with companies like Foxconn and TSMC already using Nvidia tools for factory digital twins and robotics.

South Korea's government cybersecurity agency KISA has secured access to Anthropic's Mythos AI model, a specialized cybersecurity tool, through Project Glasswing. Samsung Electronics, SK Hynix, and SK Telecom are also joining the initiative that will expand to around 150 organizations across more than 15 countries. The model helps identify and fix security vulnerabilities using frontier AI — a step toward AI-powered cyber defense at national scale.

Microsoft's Azure cloud business grew 40% last quarter as companies kept spending on AI services. But even that wasn't the ceiling—Microsoft said capacity limits in data centers held revenue back by a couple of percentage points. More significantly, Microsoft restructured its OpenAI deal, giving up exclusive rights to resell OpenAI products on Azure while keeping a 20% revenue share through 2030. OpenAI can now sell through Amazon and Google clouds. Amazon quickly responded by agreeing to invest up to $50 billion in OpenAI and expand its AWS partnership by $100 billion over eight years. The moves mark a shift in how AI power gets distributed across the biggest cloud platforms.

Goldman Sachs is reorganizing its investment banking team to focus on digital infrastructure and AI deals, signaling that big money sees AI data centers and chips as the next boom area. This means more investment flowing into AI, which could speed up new products.

After Google's I/O conference announcements about expanding AI in search, social media backlash has been intense. Critics argue Google is replacing traditional link-based search with AI chatbot features that many users never asked for. DuckDuckGo reports a 30% week-over-week jump in US installs, directly attributing it to Google's AI push. The privacy-focused search engine is positioning itself as an alternative for users wanting control over their AI exposure. Reddit users have joked that Google may drive people to switch to competitors.

Intel is pushing for AI agents to run directly on your PC instead of sending data to the cloud. At Computex 2026, Intel CEO Tan Lip-Bu showed how these 'agent computers' can handle sensitive tasks like analysing business deals without your private data leaving your device. The company showcased its Core Ultra Series 3 processor powering this privacy-first approach, with major PC makers like Lenovo, HP and Dell already selling machines with this chip. Gartner predicts AI PCs will make up over half the global PC market by 2026. Nvidia also joined the fray with its new RTX Spark chip for consumer laptops, promising AI capabilities where you simply ask the PC to do things instead of clicking through menus.

Anthropic has partnered with US cybersecurity firm Tenable to integrate its Claude AI into Tenable's security platform. The new Hexa AI engine can handle multi-step reasoning, automatically create support tickets, and generate security policies. This helps companies triage and fix cyber threats faster without manual work. It's part of a broader trend where AI labs partner with enterprise software makers to embed AI reasoning into existing tools.

Israeli cybersecurity startup Cyera just landed $540 million in fresh funding, pushing its valuation to $6 billion in just six months. The company automates finding and protecting sensitive company data — a hot need as businesses adopt AI and generate more sensitive data across more systems. Cyera's market share in data security posture management jumped from 5% to 18% in one year, showing enterprises are prioritizing this area fast.

Indian data center stocks surged this week after Prime Minister Modi met with OpenAI's Sam Altman and Anthropic's Dario Amodei in New Delhi, signaling India's push to become a global AI hub. Shares in companies like E2E Networks and Netweb Technologies rose over 18%, adding roughly $4 billion in market value. Major investments from Google, Microsoft, Reliance Industries, and the Adani Group are driving the rally.

### NEED TO KNOW * United Airlines is enhancing its AI-powered Every Flight Has a Story program to provide real-time updates on flight delays * The upgraded system aims to deliver clear, automated messages without human intervention using generative AI technology * United also uses AI for its ConnectionSaver tool, which has helped save nearly 54,000 flight connections in Denver, according to the company United Airlines is planning to upgrade its use of AI to provide information on delayed flights. During an [investor presentation](https://stockanalysis.com/stocks/ual/transcripts/669253-bernstein-42nd-annual-strategic-decisions-conference/) last week, United CEO Scott Kirby said the airline is working on upgrading its Every Flight Has a Story initiative, which uses generative AI to deliver

Young workers who followed traditional advice — study hard, stay hungry — are still struggling to find jobs as AI and automation take over routine work. The piece argues human skills like understanding context, pushing back on bad ideas, and taking responsibility can't be easily automated, and no training program can hand you these abilities.

Remark, an AI startup that creates virtual shopping assistants trained by over 60,000 human product experts, has raised $16 million in Series A funding. The company claims its AI advisors boost shopper conversion rates to 28%—far above the typical 1% industry average. Investors see huge potential in using AI to replicate what in-store experts do online, especially for high-consideration purchases like beauty and fashion. Remark says it has maintained 100% customer retention while growing revenue fourfold year-over-year. The funding will fuel expansion into new product categories and improvements to its AI tools.

Palo Alto Networks is acquiring Israeli startup Koi for around $400 million to build a new 'Agentic Endpoint Security' product line. The purchase addresses a growing gap where AI agents operating on employee devices have broad data access that bypasses traditional security tools. It's part of a broader platform war as Microsoft, CrowdStrike, and Zscaler race to build unified AI security suites.

Uber is building out its India operations with two new engineering campuses in Bengaluru and Hyderabad for nearly 10,000 people by 2027, plus its first Indian data center through a partnership with Adani Group. The move signals how tech companies are turning to local infrastructure partners to meet AI computing needs and data localization requirements in high-growth markets.

ASUS unveiled Zenni Claw at Computex 2026 in Taipei—a new AI platform designed for businesses that can be set up in just three clicks. The company also launched an AI x ESG Platform to help companies track and act on sustainability data. Both tools are aimed at enterprise customers looking to add AI capabilities without complex setup.

Singapore has climbed into OpenAI's top five markets globally for Codex usage — the AI coding tool now has 4 million weekly active users. OpenAI is working with big consulting firms like Accenture, Capgemini, and PwC to bring Codex to enterprise teams. It's a sign that AI coding assistants are gaining real traction among professionals here.

BDx Data Centers has locked in 1.2 gigawatts of electricity from Indonesia's state utility PLN—the largest power allocation held by any single data center operator in the country. That's enough juice to run roughly a million typical homes. The power will support three Indonesian sites used for AI training and cloud workloads, including facilities running Nvidia's H100 chips. The deal signals how seriously operators are competing for energy resources in Southeast Asia, where AI facilities are hungry for reliable electricity. BDx expects to consume all the capacity within two to three years as demand from Indonesian enterprises deploying GPUs grows.

United Airlines is building an AI system that will automatically explain flight delays to passengers in plain language—without any human involvement. The airline's 'Every Flight Has a Story' program already uses generative AI, but currently combines it with human editors. The goal is full automation: the AI will text customers about what's happening with their specific flight, including sending maintenance videos. United also uses AI for its ConnectionSaver tool, which decides whether to hold departing planes for connecting passengers—it has saved nearly 54,000 connections in Denver alone. CEO Scott Kirby calls this infrastructure that will help United run its entire operation more efficiently.

Elon Musk's SpaceX and Tesla are proposing a $55 billion semiconductor plant in Texas called Terafab, potentially expanding to $119 billion. The idea: build captive chip supply for AI and computing needs instead of relying on Samsung or TSMC. If it goes ahead as planned, it would rival TSMC's Arizona investment and even exceed the US CHIPS Act budget.

Nvidia has poured US$2 billion into chipmaker Marvell Technology, deepening their partnership on custom AI semiconductors. Marvell makes chips for Amazon Web Services and other hyperscalers. The deal links Marvell's optical interconnect technology with Nvidia's NVLink Fusion platform, designed to handle faster data movement inside AI data centers. Jensen Huang reportedly said Marvell could reach a US$1 trillion valuation. Alphabet and Meta are expected to spend over US$630 billion on AI infrastructure this year.

South Korea's government cybersecurity agency KISA has secured access to Anthropic's Mythos AI model, a specialized cybersecurity tool, through Project Glasswing. Samsung Electronics, SK Hynix, and SK Telecom are also joining the initiative that will expand to around 150 organizations across more than 15 countries. The model helps identify and fix security vulnerabilities using frontier AI — a step toward AI-powered cyber defense at national scale.

Microsoft's Azure cloud business grew 40% last quarter as companies kept spending on AI services. But even that wasn't the ceiling—Microsoft said capacity limits in data centers held revenue back by a couple of percentage points. More significantly, Microsoft restructured its OpenAI deal, giving up exclusive rights to resell OpenAI products on Azure while keeping a 20% revenue share through 2030. OpenAI can now sell through Amazon and Google clouds. Amazon quickly responded by agreeing to invest up to $50 billion in OpenAI and expand its AWS partnership by $100 billion over eight years. The moves mark a shift in how AI power gets distributed across the biggest cloud platforms.

Anthropic has partnered with US cybersecurity firm Tenable to integrate its Claude AI into Tenable's security platform. The new Hexa AI engine can handle multi-step reasoning, automatically create support tickets, and generate security policies. This helps companies triage and fix cyber threats faster without manual work. It's part of a broader trend where AI labs partner with enterprise software makers to embed AI reasoning into existing tools.

Israeli cybersecurity startup Cyera just landed $540 million in fresh funding, pushing its valuation to $6 billion in just six months. The company automates finding and protecting sensitive company data — a hot need as businesses adopt AI and generate more sensitive data across more systems. Cyera's market share in data security posture management jumped from 5% to 18% in one year, showing enterprises are prioritizing this area fast.

Remark, an AI startup that creates virtual shopping assistants trained by over 60,000 human product experts, has raised $16 million in Series A funding. The company claims its AI advisors boost shopper conversion rates to 28%—far above the typical 1% industry average. Investors see huge potential in using AI to replicate what in-store experts do online, especially for high-consideration purchases like beauty and fashion. Remark says it has maintained 100% customer retention while growing revenue fourfold year-over-year. The funding will fuel expansion into new product categories and improvements to its AI tools.

Nvidia is collaborating with Intel and Taiwan's MediaTek to build new laptop chips that combine CPU, GPU, and AI processing on a single chip. The chips target AI PCs and will appear in devices from Dell and Lenovo. Nvidia is also investing $5 billion in Intel stock as part of this push.

Court documents reveal Microsoft was targeting a $92 billion return on its early $13 billion OpenAI investment, with the deal granting Microsoft a 27% stake. OpenAI paid Microsoft $493.8 million in revenue share in 2024, rising to $865.8 million by late 2025. The disclosures came during Elon Musk's lawsuit against OpenAI and Microsoft over its shift to a for-profit model.

Intel is pushing for AI agents to run directly on your PC instead of sending data to the cloud. At Computex 2026, Intel CEO Tan Lip-Bu showed how these 'agent computers' can handle sensitive tasks like analysing business deals without your private data leaving your device. The company showcased its Core Ultra Series 3 processor powering this privacy-first approach, with major PC makers like Lenovo, HP and Dell already selling machines with this chip. Gartner predicts AI PCs will make up over half the global PC market by 2026. Nvidia also joined the fray with its new RTX Spark chip for consumer laptops, promising AI capabilities where you simply ask the PC to do things instead of clicking through menus.

Indian data center stocks surged this week after Prime Minister Modi met with OpenAI's Sam Altman and Anthropic's Dario Amodei in New Delhi, signaling India's push to become a global AI hub. Shares in companies like E2E Networks and Netweb Technologies rose over 18%, adding roughly $4 billion in market value. Major investments from Google, Microsoft, Reliance Industries, and the Adani Group are driving the rally.

### NEED TO KNOW * United Airlines is enhancing its AI-powered Every Flight Has a Story program to provide real-time updates on flight delays * The upgraded system aims to deliver clear, automated messages without human intervention using generative AI technology * United also uses AI for its ConnectionSaver tool, which has helped save nearly 54,000 flight connections in Denver, according to the company United Airlines is planning to upgrade its use of AI to provide information on delayed flights. During an [investor presentation](https://stockanalysis.com/stocks/ual/transcripts/669253-bernstein-42nd-annual-strategic-decisions-conference/) last week, United CEO Scott Kirby said the airline is working on upgrading its Every Flight Has a Story initiative, which uses generative AI to deliver

Vietnam has become a key hub in Nvidia's sovereign AI strategy. Local firm FPT built a population-scale AI training dataset for Vietnam, while Viettel is developing a national legal AI app using Nvidia's models. Singapore is among the countries included in Nvidia's sovereign AI network, alongside South Korea, India, Japan, and others.

Your colleagues might be secretly using AI tools at work that could be leaking company secrets. A new workplace trend called "Shadow AI" is emerging across Asia, where employees download AI chatbots and automation tools without IT approval. In Singapore, 68% of workers use AI frequently on the job, but only 14% stick to company-approved tools. This means sensitive data from your workplace emails, documents, or customer records could be feeding into public AI systems without your boss's knowledge. Companies are now turning to DNS monitoring and clear AI usage policies to catch these "shadow employees" before they accidentally expose private information.

Hyundai is rolling out Boston Dynamics Atlas humanoid robots across its car plants, targeting 30,000 robots. The company set up new units for software-defined factories and trade strategy amid US tariff pressures. Atlas handles material handling and parts sorting, with first real-world deployment at a Georgia plant from 2028.

Intel shares surged 58% over nine consecutive days, the chipmaker's best run since the 1970s. The rally was driven by partnerships with Google (Xeon 6 CPUs for AI), Nvidia (US$5B investment and custom chip development), and Elon Musk's Terafab project in Austin for SpaceX, xAI, and Tesla. Intel also bought back its Ireland chip plant for US$14.2 billion.

Young workers who followed traditional advice — study hard, stay hungry — are still struggling to find jobs as AI and automation take over routine work. The piece argues human skills like understanding context, pushing back on bad ideas, and taking responsibility can't be easily automated, and no training program can hand you these abilities.

BDx, a Singapore-based data center developer, has locked in 1.2 gigawatts of power capacity in Indonesia through a deal with state utility PLN—the biggest power commitment ever secured by a data center operator in Indonesia. The agreement covers three facilities including the CGK4 AI Campus in West Java, which is Indonesia's first NVIDIA DGX-Ready-certified site built for AI training with H100-class GPUs. A second site in South Jakarta will support liquid-cooled, high-density AI compute including the next-generation NVIDIA GB200. For everyday users, this matters because AI services—from chatbots to cloud features—run on data centers that need enormous, reliable power. This deal signals that Southeast Asia is becoming a key battleground for AI infrastructure.

OpenAI, Anthropic, and xAI are all actively hiring in Singapore, with OpenAI leading with 22 open positions at its Singapore office. The city-state has become a key base for frontier AI labs including Google DeepMind and AMI Labs, reflecting Singapore's ambition to be Asia's AI hub with nearly US$780 million committed to AI research over four years.

Marketing agencies are rethinking their approach as AI makes generic content cheap. Instead of acting as 'order-takers' who execute briefs blindly, the best partners now co-create strategy with clients from the start. Tech in Asia Studios describes its own model as an 'ecosystem interpreter' that embeds teams within client strategy rather than working from outside.

TSMC plans to build advanced chip packaging facilities in Arizona by 2029, easing bottlenecks in US AI chip production. The facilities will support CoWoS and 3D-IC packaging used in AI chips for Nvidia and others. TSMC has a deal with Amkor to provide packaging and test services for wafers made at its Phoenix plants. This marks a shift toward a more self-sufficient US chip supply chain.

After Google's I/O conference announcements about expanding AI in search, social media backlash has been intense. Critics argue Google is replacing traditional link-based search with AI chatbot features that many users never asked for. DuckDuckGo reports a 30% week-over-week jump in US installs, directly attributing it to Google's AI push. The privacy-focused search engine is positioning itself as an alternative for users wanting control over their AI exposure. Reddit users have joked that Google may drive people to switch to competitors.

UK publishers can now choose not to appear in Google's AI search summaries, thanks to a new rule from the Competition and Markets Authority. The regulator says this will give publishers more power to negotiate content deals with Google. Many websites have seen big drops in traffic since Google started placing AI summaries at the top of search results. Google controls over 90% of UK search and has nine months to implement the changes, which also require proper attribution when content appears in AI results.

Nvidia is building a new campus called Constellation in Taipei, spanning nearly four hectares, with construction starting in the next couple of months. CEO Jensen Huang unveiled the plan at an employee meeting on May 28. The company already has an office in Kaohsiung and is hiring heavily in Taipei as its Taiwan supply chain grows. Huang said Nvidia sees growth in agentic AI and that physical AI will change manufacturing in Taiwan, with companies like Foxconn and TSMC already using Nvidia tools for factory digital twins and robotics.

A new study finds that the Asia-Pacific region leads the world in AI adoption but is struggling to find enough skilled workers. For everyday users, this means companies may rush to roll out AI tools, but quality and support could suffer without enough talent.

AI company Anthropic agreed to pay Elon Musk's xAI $1.25 billion per month through 2029 for computing power from its data center. This massive deal shows how AI firms are scrambling for hardware and could lead to faster, more reliable AI services for users.

A US trade agency launched a patent investigation against memory chip makers SK hynix and Kioxia over 3D chip designs. If imports are banned, it could disrupt the supply of memory chips used in AI servers, potentially slowing down AI development and raising costs.

Goldman Sachs is reorganizing its investment banking team to focus on digital infrastructure and AI deals, signaling that big money sees AI data centers and chips as the next boom area. This means more investment flowing into AI, which could speed up new products.

Microsoft has launched its own AI models to compete directly with OpenAI and Anthropic. The move marks a major shift as Microsoft — a key backer of OpenAI — develops independent AI capabilities.

Alibaba just dropped preview versions of its next Qwen AI models ahead of a big launch in Hangzhou. Early tests rank them among the world's top AI systems, beating other Chinese labs but still behind OpenAI, Google, and Anthropic. The company is burning billions in profit for market share — free cash flow dropped to a $2.5 billion outflow last quarter — but AI product revenue has grown at triple-digit rates for 11 straight quarters.

Taiwan is reportedly exploring armed robot dogs for military defense against potential threats from China. The story taps into a broader global trend of nations deploying AI-powered autonomous systems for combat and surveillance roles.

Vietnam's biggest conglomerate Vingroup just unveiled humanoid robots at two major global tech events — ICRA 2026 in Vienna and COMPUTEX Taipei. Two subsidiaries built separate models: VinRobotics showed the VR-H3, a third-generation industrial robot with remote operation via VR headset, while VinDynamics launched Dyno, a household and security assistant. Both companies built everything from scratch — mechanical parts, AI, power systems, and actuators — marking a rare Vietnamese tech debut on the world stage.

# Google unveils Gemini AI for multi-step Android tasks [News](https://www.techinasia.com/)Premium * [News](https://www.techinasia.com/) * [Jobs](https://www.techinasia.com/jobs/search) * [Companies](https://www.techinasia.com/companies) * [Events](https://www.techinasia.com/events/live) * [About](https://www.techinasia.com/about) * [advertise](https://www.techinasia.com/advertise?utm_source=techinasia.com&utm_medium=website&utm_campaign=intl-navbar) * [Asia Economic Summit](https://aes.techinasia.com/?utm_source=navbar-tia-homepage&utm_medium=website&utm_campaign=events-flagship-aes) * [Sign Up](https://www.techinasia.com/auth) * [Log In](https://www.techinasia.com/auth?ref=nav_login) [Subscribe](https://www.techinasia.com/subscription) [I'm fundraising](https://www.techinasia.com/compani

SaaStr, a B2B software community, built custom AI tools to handle the boring sales tasks humans skip. The trick: start with ONE annoying workflow, connect it to your existing database like Salesforce, and let the tool grow as people actually use it. They booked 614 meetings worth $85K each by automating scheduling links for leads humans deemed not worth their time. The warning: AI can repeat mistakes thousands of times before anyone notices, so human oversight is non-negotiable.

Nvidia announced RTX Spark, a new processor for Windows laptops, at GTC Taipei during Computex. The company also detailed Vera and Vera Rubin chips designed for AI data centers, alongside infrastructure updates. RTX Spark brings AI capabilities directly to consumer Windows PCs.

Researchers have created AI-powered computer worms that can spread between systems without human interaction. Unlike traditional viruses that need users to click something, these worms autonomously infect devices, copy data, and jam communications. Security experts are calling this a paradigm shift in cyber threats.

Russia's Sberbank is offering its AI model to developing nations, positioning it as a tool to close the digital gap. This is part of a broader geopolitical competition over AI influence—Western and Chinese AI dominate, and Russia is now pitching its technology as an alternative for countries seeking more options.

Enterprise companies are now being judged on how efficiently they run AI workloads as data center energy consumption surges. The focus is shifting beyond infrastructure to the logistics and systems that support AI computing, with pressure to scale without increasing carbon emissions.

This is a general discussion piece exploring how AI is being used in sports broadcasts, gaming, music, and events. It covers things like AI-generated commentary, personalized highlights, and virtual reality experiences. Nothing particularly new or groundbreaking.

Google is rolling out AI-powered voice scam detection for Android phones before the end of June. The feature uses AI to analyze calls in real-time and alerts users when suspicious patterns are detected. This builds on existing scam call protections already available on Android devices.

Singapore's Urban Redevelopment Authority is showcasing practical AI tools for city planning at a new exhibition. One standout project uses sensors on seniors' bodies and robots to identify environmental hazards that could cause falls — helping planners design safer neighbourhoods for an ageing population. Other tools include a handheld scanner that detects building defects in 30 seconds using AI, and cameras that flag illegal dumping and unsafe parking in real time.

MedPlanner, a health tech company based in Kuala Lumpur, has launched two AI products: HEMI Health, an AI medical scribe with legal protections for doctors, and AskHEMI, a multilingual app helping patients manage chronic conditions. Both run on MedPlanner's proprietary health AI system.

Asus is using Nvidia's DSX platform to create 'digital twins' of data center setups—virtual simulations of power, cooling, networking, and storage—before building physical infrastructure. This helps companies deploy AI data centers faster and with fewer costly mistakes.

Meta is rolling out AI-powered 'business agents' designed to handle routine enterprise tasks like customer service, scheduling, and internal communications. These tools aim to automate daily operations, potentially reducing workloads for small business teams.

Taiwan's Asus showcased its 'AI City' concept and new laptops powered by Nvidia chips at Computex, a major tech showcase. The AI City appears to integrate AI into urban environments, while the laptops target creative professionals and gamers needing heavy AI processing power.

The US FDA is evaluating an AI tool that predicts liver injury risk from medications before they reach patients. If approved, doctors could use it to flag potentially dangerous drug interactions early, improving drug safety testing.

Singapore organisations rank among Asia Pacific's most advanced on AI governance policy, but a new JFrog report exposes consistent failure to translate those policies into actual enforcement. This leaves enterprises exposed, particularly around software supply chain security. The gap between having AI policies and actually enforcing them is a real risk for businesses.

UN researchers warn that data centres powering AI will double their electricity and water use by 2030. Last year they consumed 448 terawatt-hours of power (more than Saudi Arabia) and 4.5 trillion litres of water. Water use alone is set to reach 9.3 trillion litres. The AI tools you use daily run on real physical infrastructure with significant environmental costs.

Major AI chatbots like ChatGPT and Microsoft Copilot are spreading pro-Russian propaganda as if it were verified fact, according to a NewsGuard investigation. The problem is worse in non-English languages. Chatbots tend to repeat whatever information is most widespread online, not what's most reliable. Researchers warn this could get worse as more people use AI to get their news.

Cohesity, an AI-powered data security company, has patented a new RAG (retrieval-augmented generation) architecture specifically designed for backup systems. This is a technical patent for enterprise data security platforms.

Billionaire investor Ray Dalio breaks down what he sees as fundamentally different philosophies driving AI development in China versus the United States. China's approach tends toward state coordination, while the US leans on private market competition. Dalio suggests this ideological clash will be a major driver of geopolitical and economic outcomes in the decade ahead.

A global shortage of memory chips is hitting automakers and retailers, driving up prices across the board. Since memory chips are essential for AI systems and everyday electronics alike, this supply crunch could eventually push up costs for AI hardware, consumer devices, and new cars.

A leaked internal Microsoft document reveals the company explicitly wants to make people 'addicted' to its new AI assistant 'Scout.' The three-phase strategy reportedly starts with hooking users, then building dependency so they use it daily. Over 1,000 Microsoft employees, including CEO Satya Nadella, are already using the tool. AI researchers have long warned about mental health risks from addictive chatbot design, and this document appears to admit that's exactly the goal.

Tencent Cloud has partnered with retail solutions provider UXSoft Global Limited to deliver a large-scale smart retail transformation for Hung Fook Tong, Hong Kong’s largest Chinese herbal product retailer by store count, migrating the brand’s mission-critical point-of-sale systems across more than 100 retail outlets to the cloud. The partnership was announced at Tencent Cloud Day […] The post Tencent Cloud Powers Hung Fook Tong’s Smart Retail Overhaul appeared first on techcoffeehouse.com .

OpenAI's CFO, Sarah Friar, confirmed the company is building an AI hardware device with legendary iPhone designer Jony Ive, and it will be unveiled before the year ends. She's tried it herself and calls it 'lovable' and 'natural.' She refused to confirm — or deny — reports it's an earpiece, joking that Ive would 'steal her teenage son' if she spilled. OpenAI bought Ive's startup IO for $6.5 billion last year.

Sophos released real-world data from its AI-powered security operations centre. The system now fixes threats in 89 seconds on average, and AI handles over half of all detected cases without a human touching them. This shows how automation is reshaping cybersecurity—faster responses, fewer manual interventions needed.