US Data Center Investment Could Exceed $880 Billion by 2027

AI investment is expanding rapidly, but questions about returns and leadership risk are pushing investors to look beyond Nvidia and hardware alone. A Federal Reserve Board working paper published November 12, 2026, says announced US data-center investment plans rose from about $300 billion in 2024 to more than $1 trillion in 2025. Annual investment could exceed $880 billion by the end of 2027, above 2% of US GDP, yet productivity and employment gains remain modest. About half of the high-tech equipment spending is imported, and key hardware prices are rising nearly 8% annually. Investors cited in the articles expect AI spending to remain strong over the next two to three years, while warning that power demands, politics and weak cash-flow returns could create volatility.

That uncertainty is prompting attention to companies across the AI supply chain. Beyond Nvidia, chip-related names include Nanya Technology, which says AI infrastructure and server products account for over 20% of revenue and targets 40%–50% by 2029–30; Kulicke and Soffa, which plans to expand Singapore capacity to support about US$400 million in Advanced Solutions revenue by the first half of fiscal 2027; and inspection, testing and memory suppliers Camtek, Aehr Test Systems, Rambus and CEVA. Broadcom, Applied Materials and ASML also feature in coverage of AI hardware demand. Elsewhere, Benchmark Electronics reported AI- and cloud-driven AC&C revenue growth of 41% in Q1 and 71% in Q2 2026, while Avnet estimates AI and data-center work contributes 10%–15% of revenue. Cooling and power suppliers are also seeing demand: Carrier cites a roughly US$2 billion 2026 data-center backlog, and Vertiv reports a US$15 billion backlog and 252% year-over-year growth in Q4 2025 organic orders.

Major technology companies are broadening their offerings and supplier networks. Microsoft has added Cerebras hardware and software to support future AI workloads, alongside Nvidia and its own Maia and Cobalt chips; its planned AI and data-center spending is about US$175 billion–US$190 billion in calendar 2026. Intel and RadixArk are optimizing the open-source SGLang inference framework for Intel Xeon CPUs and accelerators, including Intel AMX, memory use and coordination across processors and accelerators. They plan to contribute changes publicly. Salesforce’s Agentforce brings autonomous agents to its customer platform, which generates US$43.9 billion in revenue; Oracle provides cloud infrastructure and generative AI services, and OpenAI expanded its cloud use to Oracle Cloud Infrastructure in 2024. Palantir’s Artificial Intelligence Platform supports AI use in government and commercial operations. The articles also identify profitable AI-linked firms including Broadcom and Japan’s Infomart.

Applications beyond chips are drawing interest, including healthcare, business software and financial services. The articles profile AI-focused healthcare firms in Australia, Japan, the UK and elsewhere: examples include Artrya’s Salix coronary-CT risk assessment software, Renalytix’s KidneyIntelX risk scores, and AstraZeneca’s oncology AI work with Tempus and Pathos. Haleon says AI tools improved forecast accuracy by 5%–6%. Freshworks has more than 7,000 paying Freddy AI customers, while Data#3 helps organizations deploy Microsoft 365 Copilot. Dicker Data expects AI-related revenue of A$100 million–A$150 million in FY26. The coverage also highlights smaller firms such as Intapp and HeartFlow, as well as ExaU, Exaforce’s service for building collaborative security agents; Exaforce raised a US$125 million Series B. Other applications range from AI-assisted insurance claims to an early-stage tool designed to flag property-sale problems.

Businesses and public institutions are weighing the benefits against operational and social risks. UK banks including HSBC, Barclays, Lloyds, NatWest and Nationwide restrict or block third-party agents from accessing customer accounts, especially where transactions are possible. A Bank of England and FCA survey found 80% of responding firms had deployed AI agents, but just 2% used agents with broad autonomy; firms cited security, third-party dependence and hallucinations. A report on data-center growth calls for legally binding Community Benefit Agreements to give rural communities a say in jobs, education and environmental protections. In sports, a slower AI rollout could give organizations time to improve data, security and oversight; NASCAR says in-house AI has replaced US$20,000–US$30,000 a year in outside digital tools. A November 19, 2026, summit in Cincinnati will bring AI and data leaders together to discuss enterprise strategy and governance.

For Malaysia, the stakes include both opportunity and exposure: electrical and electronics exports rose 49.5% year over year to RM766.38 billion in the first nine months of 2025, or 48.8% of total exports. Analysts say a valuation correction would be easier to absorb than a sharp cut in AI capital spending, which could affect technology shares, exports and data-center projects. Investors are also being encouraged to consider AI applications and biotechnology, rather than focusing only on local infrastructure and semiconductor companies. Separately, articles highlight firms tied to trading activity amid Nvidia leadership concerns and small-cap AI companies in the US, Australia and Japan; one article describes Jensen Huang’s abrupt exit from Nvidia as a shock to investor confidence.

Key Takeaways

  • The Federal Reserve paper estimates US data-center investment could exceed US$880 billion annually by the end of 2027, while broad productivity and employment effects remain modest.
  • Microsoft adds Cerebras to its AI chip supplier network; planned AI and data-center spending is US$175 billion–US$190 billion in calendar 2026.
  • Intel and RadixArk are optimizing open-source SGLang for Xeon CPUs and accelerators, with plans to contribute their changes to the public repository.
  • Salesforce’s Agentforce adds autonomous AI agents to its customer platform; OpenAI expanded its use of Oracle Cloud Infrastructure in 2024, while Palantir markets its AI Platform for government and commercial operations.
  • AI hardware demand spans suppliers beyond Nvidia, including Nanya Technology, Kulicke and Soffa, Camtek, Broadcom, Applied Materials, ASML, Aehr Test Systems, Rambus and CEVA.
  • Carrier reports a roughly US$2 billion 2026 data-center backlog, while Vertiv reports a US$15 billion backlog and 252% year-over-year Q4 2025 organic order growth.
  • Healthcare and business applications include Artrya’s Salix, Renalytix’s KidneyIntelX, AstraZeneca’s oncology AI work, Freshworks’ 7,000-plus paying Freddy AI customers, and Data#3’s Microsoft 365 Copilot deployments.
  • UK banks limit third-party AI agents’ account access; a Bank of England and FCA survey found 80% of respondents had deployed agents, but only 2% used broadly autonomous agents.
  • Malaysia’s electrical and electronics exports reached RM766.38 billion in the first nine months of 2025, up 49.5% year over year; analysts warn a sharp AI-capital-spending slowdown would pose risks.
  • Exaforce launched ExaU for building collaborative security agents and raised US$125 million in a Series B; a November 19, 2026, Cincinnati summit will focus on AI and data leadership.

Three AI Small Caps to Watch as Credit Risks Rise

As refinancing costs climb, investors are questioning the risks tied to large AI companies, creating interest in smaller AI firms. Freshworks, Intapp and HeartFlow are three companies highlighted from a screen of US stocks. Freshworks sells customer and employee software, including Freshdesk Omni and Freshchat, and has more than 7,000 paying Freddy AI customers. Intapp makes cloud software for professional and financial firms, while HeartFlow uses AI to turn CT scans into 3D heart models that help assess coronary artery disease. The article says the full screen found 39 other AI focused small cap stocks.

Three Australian AI Healthcare Stocks to Watch

The article highlights ImExHS, Singular Health Group and Artrya as Australian companies applying AI to healthcare. Sydney based ImExHS supplies cloud medical imaging software and radiology services, with about A$31 million in revenue concentrated in Latin America. Singular Health Group turns CT, MRI and PET scans into 3D imaging and file sharing tools, while its 3DICOM software generated about A$1.9 million. Artrya’s Salix software uses AI to assess heart attack risk from coronary CT scans. The article says the full screen found two other AI healthcare companies.

Three Japanese AI Stocks to Watch in November 2026

The article names ExaWizards, Finatext Holdings and Sinfonia Technology as Japanese AI focused companies to watch. ExaWizards develops AI tools for automation, predictive maintenance and data based decisions. Finatext provides cloud financial technology, AI data analysis and finance apps, with revenue from financial infrastructure, big data analysis and fintech services. Sinfonia makes automation equipment for factories, semiconductors, aerospace and laboratories, including AI ready cell culture systems. The article says its full small cap screen found 10 other companies.

Three Tech Stocks in Focus After Nvidia Leadership Shock

Jensen Huang’s abrupt exit from Nvidia has led investors to reconsider the risks of relying on a single leader and governance structure. The article highlights Samsung Electronics, Keysight Technologies and Delta Electronics (Thailand) as large technology companies with AI or cloud exposure. Samsung makes semiconductors and devices, while Keysight supplies testing tools for semiconductors, cloud infrastructure and wireless networks. Delta Electronics (Thailand) sells power electronics and infrastructure hardware, including data center equipment and AI cooling solutions. The full screen identified 29 other large technology companies.

Three Stocks to Watch as AI Volatility Drives Trading Activity

Volatility linked to Nvidia’s leadership crisis has increased attention on exchanges and financial platforms that may benefit from higher trading activity. Tel-Aviv Stock Exchange operates Israel’s main equities and derivatives market, and its move to a Monday through Friday trading week has been linked to increased foreign activity and average daily volumes. Anand Rathi Wealth serves high and ultra high net worth clients in India, while Prudent Advisory Services distributes mutual funds, insurance and other financial products. The article notes that changes in client activity and product fees could affect how trading swings translate into revenue. Its full screen found 37 other exchanges and trading platforms.

Three Australian AI Stocks Worth Watching

A November 2026 article highlights Pureprofile, Dicker Data and Data#3 as Australian companies with links to AI. Pureprofile uses AI and automation in platforms such as Datarubico, Insights Creator and Sample Only. Dicker Data distributes AI hardware and software, and expects AI related revenue of A$100 million to A$150 million in FY26. Data#3 helps organizations deploy AI tools such as Microsoft 365 Copilot, alongside its cloud and security services.

Three AI Chip Stocks Beyond Nvidia to Watch

The article looks at Nanya Technology, Kulicke and Soffa Industries, and Camtek as companies tied to AI chip production and supply. Nanya makes DRAM memory, and says AI infrastructure and server products already account for more than 20% of its revenue, with a target of 40% to 50% by 2029 to 2030. Kulicke and Soffa makes chip assembly and advanced packaging equipment, and plans to expand Advanced Solutions capacity in Singapore to support about US$400 million in segment revenue by the first half of fiscal 2027. Camtek supplies inspection and measurement tools for chip and memory production.

Three Profitable AI Stocks to Watch in November 2026

The article highlights Broadcom, Japan's Infomart and Palantir Technologies as AI focused companies that are generating earnings. Broadcom sells data center chips and private cloud software, with revenue from Semiconductor Solutions and Infrastructure Software. Infomart uses AI-OCR to automate business workflows and reported nine month revenue of ÂĄ15,253.68 million and net income of ÂĄ1,973.2 million, though its P/E ratio was 83.2. Palantir builds AI data platforms for government and commercial customers, and its Artificial Intelligence Platform supports the use of AI in business operations.

Three Semiconductor Stocks Positioned for AI Growth

The article examines Broadcom, Applied Materials and ASML Holding as companies linked to AI hardware spending. Broadcom supplies chips and software for digital infrastructure and is expanding its custom silicon and networking businesses. Applied Materials builds equipment used to make advanced wafers, and its Semiconductor Systems business generated about US$26.1 billion in revenue. ASML makes lithography tools that help foundries print tiny chip features, and an investor cited strong demand for its machines through at least 2028, with some customers seeking orders for 2029 and 2030.

Three AI Chip Stocks With Less Reported China Exposure

After Nvidia shares fell 28% from their peak, the article points to Aehr Test Systems, Rambus and CEVA as AI hardware companies with lower reported China exposure. Aehr makes burn-in and reliability testing equipment for high performance chips, including its FOX wafer-level and Sonoma package-level platforms. Rambus supplies memory interface chips and security technology, and the article describes a potential US$600 million MRDIMM market around 2028, with meaningful Rambus revenue expected in 2027. CEVA licenses digital signal processing, AI and wireless technology, with new Edge AI licensing deals that could lead to higher royalties as products reach mass production over the next 18 to 24 months.

Microsoft adds Cerebras to its AI chip supplier network

Microsoft has formed an AI partnership with Cerebras, adding another specialist chip supplier to its infrastructure. The deal will bring Cerebras hardware and software into Microsoft’s systems to support future AI workloads, broadening its chip options beyond Nvidia and its own Maia and Cobalt chips. Microsoft’s planned AI and data center spending is about US$175 billion to US$190 billion in calendar 2026. The article says upcoming updates on AI revenue and chip use will help show whether the added capacity is being used efficiently.

Three supply chain stocks with ties to AI hardware

The article highlights Benchmark Electronics, Sanmina and Avnet as companies with roles in AI hardware supply chains, where tracking and traceability are under greater scrutiny. Benchmark designs, builds and tests complex hardware, and reported year-over-year AC&C revenue growth of 41% in Q1 and 71% in Q2 2026, driven by AI and cloud programs. Sanmina provides electronics manufacturing, logistics and repair services, and is expanding production capabilities for AI racks and cloud systems. Avnet distributes electronic components and embedded solutions worldwide, with management estimating that AI and data center work makes up about 10% to 15% of revenue, mostly in Asia and Taiwan.

Three AI stocks with earnings growth above 45%

The article features Salesforce, Oracle and Super Micro Computer as companies on an undervalued AI stock shortlist. Salesforce’s Agentforce suite brings autonomous AI agents to its cloud-based customer platform, which generates US$43.9 billion in revenue and has a market value of about US$179.3 billion. Oracle’s cloud infrastructure and generative AI services serve large enterprises, and OpenAI expanded its cloud use to Oracle Cloud Infrastructure in 2024. Super Micro Computer makes servers and storage for AI, cloud and data centers, with about US$54.0 billion in revenue from its server solutions segment and a market value near US$17.7 billion.

Three US stocks powering AI data centers with cooling and energy systems

The article looks at Carrier Global, Vertiv Holdings and SPX Technologies, which supply cooling and other systems used in data centers and critical infrastructure. Carrier Global provides HVAC and energy management systems, and its 2026 data center revenue backlog is about US$2 billion; it also plans a US manufacturing site to roughly double its Americas data center capacity. Vertiv supplies power and cooling equipment, including liquid cooling, and had a US$15 billion backlog; its Q4 2025 organic orders rose 252% year over year. SPX Technologies makes HVAC cooling systems and detection and measurement products, with most of its sales in the United States.

Three British healthcare stocks using AI

The article covers Renalytix, Haleon and AstraZeneca as UK-listed companies using AI in healthcare and medicine. Renalytix’s KidneyIntelX platform uses biomarkers, genetics and electronic health records to create kidney disease risk scores that support clinical decisions, research and drug discovery. Haleon sells consumer healthcare products and uses a Microsoft AI partnership to support product development and data-driven marketing; its AI tools improved forecast accuracy by 5% to 6%. AstraZeneca is working with Tempus and Pathos on an oncology AI model aimed at improving cancer diagnosis and treatment choices.

Investors look beyond AI hardware to applications and biotech

Investors are increasingly asking whether heavy spending on AI infrastructure will lead to lasting earnings and cash flow. Franklin Equity’s Asia-Pacific client portfolio management head, William Chang, sees opportunity in practical AI products, including business software and consumer apps, as well as biotechnology tools for drug discovery. He expects global AI spending to remain strong over the next two to three years, but warns that political developments and electricity use could cause volatility. Chang says Malaysian investors can look beyond local infrastructure and semiconductor companies to opportunities across the global AI value chain.

AI spending boom faces questions over returns and risks to Malaysia

Investors are questioning whether spending on AI data centres and semiconductors will earn enough to justify its cost, and technology stocks have fallen from recent highs. In Malaysia, electrical and electronics exports rose 49.5% year on year to RM766.38 billion in the first nine months of 2025, making up 48.8% of total exports. SPI Asset Management’s Stephen Innes says the AI boom has not burst, but investors are demanding stronger cash flow and returns. A slowdown could affect Malaysian technology shares, exports, data centre projects, related industries and the ringgit, although he says an investment valuation correction would be easier for Malaysia to handle than a sharp cut in AI capital spending.

Federal Reserve paper finds huge AI data center spending but modest economic gains

A Federal Reserve Board working paper published on November 12, 2026, reports that announced US data center investment plans grew from about $300 billion in 2024 to more than $1 trillion by 2025. The paper estimates annual investment will rise from $100 billion in 2024 to more than $880 billion by the end of 2027, above 2% of US GDP. About two thirds of the spending goes to high tech equipment, and roughly half of that equipment is imported, limiting its direct contribution to domestic output. The paper also reports nearly 8% annualized price increases for key hardware and notes that economy-wide productivity and employment effects remain muted.

Intel and RadixArk work to speed SGLang AI inference on Intel hardware

Intel and RadixArk announced an engineering partnership on November 12, 2026, to optimize the open-source SGLang AI inference framework for Intel Xeon CPUs and accelerators. Their work includes tuning software for Intel AMX, improving memory use and enabling new open-weight models to run on Intel platforms from their release. The companies say they will contribute their software changes to the public SGLang repository rather than create a closed version. The collaboration also aims to coordinate workloads between Xeon processors and accelerators.

A slower AI boom could give sports organizations time to prepare

Recent shocks in the AI industry have raised questions about how quickly sports teams and leagues should adopt the technology. A slowdown could give them time to improve data, security and oversight, train staff, and choose AI tools that provide lasting value. Sports organizations are already using AI for business operations, while NASCAR has replaced $20,000 to $30,000 a year in outside digital tools with in-house AI solutions. Higher AI costs could force teams to focus on worthwhile uses, but a weaker AI market could also reduce sponsorship spending and interest from wealthy buyers looking to own teams.

Agentic AI could streamline complex property and casualty claims

Property and casualty insurers receive claims through many kinds of records, including ACORD notices, claimant accounts, damage photos, repair estimates, emails and call records. Property claims can also include aerial images, contractor reports and handwritten notes, while disasters create tighter deadlines. Insurers want to speed up processing, reduce manual data entry, and limit losses from errors and fraud. The article says older, step by step automation struggles with changing processes and unstructured evidence, while specialized AI can help, but designing systems that handle exceptions without more manual work remains difficult.

New AI tool aims to spot property sale problems early

Researchers at a leading university are developing a machine learning tool to flag possible problems in property sales before conveyancing begins. The tool is still in its early stages, and developers say it could reduce failed sales and make conveyancing more efficient. The team is working with industry experts and testing the tool to make sure it meets the industry’s needs. The developers expect to launch it in the coming months for property professionals and conveyancers.

Exaforce lets security teams build personal AI agent groups

Exaforce launched ExaU, a service that lets security workers create AI agents by describing their jobs in plain language or choosing templates. The agents keep track of past decisions and instructions, work out what to do next, and share findings, questions and recommendations. Teams can share agents, and one assessment can combine results from tasks such as threat hunting, activity audits and employee outreach. ExaU runs on customers’ data within Exaforce and uses the company’s security knowledge graph; actions affecting people, systems or communications need approval and appear in an audit log. Investors in Exaforce’s $125 million Series B round included HarbourVest Partners, Peak XV Partners, Mayfield Fund, Khosla Ventures and Seligman Ventures.

UK banks restrict AI agents amid security and competition worries

HSBC, Barclays, Lloyds, NatWest and Nationwide have blocked or limited third-party AI agents from accessing customer accounts, especially when agents could make transactions without customers. Banks are concerned about mishandled money and responsibility, while agents could also make it easier for customers to find better deals elsewhere. Apollo economist Torsten Slok warned that agents moving large amounts of household cash could cause an “agentic bank run,” and JPMorgan Chase chief executive Jamie Dimon warned against unrestricted access. Emirates NBD strategist Neeraj Makin expects agents to bring around-the-clock banking and near-instant payments by 2036, while the bank’s Leap framework is designed to govern agent access and actions. A Bank of England and FCA survey found that 80% of responding firms had deployed AI agents, but only 2% used agents with broad autonomy; firms cited data security, third-party dependencies and hallucinations as concerns.

Report urges community agreements to protect Black rural workers as AI grows

A Joint Center report says AI data centers are growing fastest in the South, which also has the largest concentration of Black rural workers. Researchers interviewed professionals in higher education, workforce development and technology, mainly in Alabama, where African Americans make up more than 40 percent of the rural population. The interviews found interest in AI, but also concerns about environmental and economic effects, limited rural input, and AI-driven land purchases that could weaken community ties. The report recommends legally binding Community Benefit Agreements between communities and private companies to set fair rules for jobs, education, the environment and enforcement. It also calls for rural communities to take part in AI development, reskilling and the creation of large language models.

Cincinnati Summit to Explore AI and Data Leadership

The Leadership Summit will take place in person at the Westin Cincinnati Downtown on November 19, 2026. It will bring together AI and data leaders to discuss strategies, governance, and innovations that support enterprise performance. Attendees can connect, learn, and gain actionable insights.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI investment data centers AI infrastructure semiconductors Nvidia AI hardware supply chain AI chips cloud computing generative AI AI agents enterprise AI AI software healthcare AI AI power demand AI productivity AI governance AI security Malaysia exports AI investment risk AI startups

Comments

Loading...