{"schemaVersion":1,"updatedAt":"2026-08-10","unit":"USD billions unless otherwise stated","combinedGuidance":{"lowBillions":695,"highBillions":720,"midpointBillions":707.5},"caveats":["The guidance is a capital-spending proxy, not a comparable audited AI-only total.","Company definitions differ across cash purchases, finance leases, and reporting periods.","Amazon guidance is companywide and includes material non-AWS investment.","Forward-looking guidance can change after the dataset review date."],"companies":[{"slug":"microsoft","company":"Microsoft","ticker":"MSFT","searchStatus":"reviewed","updatedAt":"2026-08-10","title":"Microsoft AI Capex in 2026: $190B Spending Guide","description":"Track Microsoft's roughly $190 billion 2026 capex plan, quarterly infrastructure spending, GPU and CPU mix, finance leases, and the signals investors should monitor.","keywords":["Microsoft AI capex 2026","Microsoft capital expenditure 2026","Microsoft data center spending","MSFT AI infrastructure investment"],"guidance":{"period":"Calendar 2026","display":"About $190B","lowBillions":190,"highBillions":190,"midpointBillions":190},"latestSignal":{"label":"FY2026 Q3 capex","value":"$31.9B","note":"Roughly two-thirds short-lived assets, primarily GPUs and CPUs"},"summary":"Microsoft expects roughly $190 billion of calendar-year 2026 capital expenditure. Management says approximately $25 billion of that envelope reflects higher component pricing, while demand for Azure capacity remains above available supply. The number is a company capital-spending plan, not a disclosed pure-AI total.","accountingScope":"Microsoft's reported capex includes additions obtained through finance leases as well as purchased property and equipment. That makes total capex different from cash paid for property and equipment in a quarter. The company also reports on a June fiscal year while the $190 billion guidance is explicitly framed on a calendar-year basis, so fiscal-year comparisons require normalization.","allocation":["Short-lived assets such as GPUs and CPUs that can enter service and support revenue relatively quickly.","Long-lived data-center sites and infrastructure intended to support monetization for 15 years or more.","Storage, networking and replacement equipment needed to keep Azure and first-party products operating."],"investorRead":"Microsoft provides the clearest split between short-lived compute and long-lived infrastructure among the four companies in this tracker. That split matters because servers generally begin affecting depreciation and capacity sooner than a large campus. The central question is how quickly contracted cloud demand and product usage convert the new capacity into revenue relative to the depreciation and operating costs arriving behind it.","watchItems":["Quarterly capex and the proportion allocated to short-lived compute versus long-lived sites.","Finance-lease additions versus cash paid for property and equipment.","Azure capacity constraints, backlog conversion and management's timing for bringing capacity online.","Depreciation growth and cloud gross-margin pressure as recent investments enter service."],"limitations":["The company does not disclose a standalone audited AI-capex figure.","Calendar-year guidance does not align directly with Microsoft's June fiscal year.","Finance-lease timing can make total capex volatile and different from cash investment in the same quarter."],"sources":[{"label":"Microsoft FY2026 Q3 earnings call","href":"https://www.microsoft.com/en-us/investor/events/fy-2026/earnings-fy-2026-q3","publishedAt":"2026-04-29"},{"label":"Microsoft FY2026 Q2 earnings call","href":"https://www.microsoft.com/en-us/investor/events/fy-2026/earnings-fy-2026-q2","publishedAt":"2026-01-28"}]},{"slug":"amazon","company":"Amazon","ticker":"AMZN","searchStatus":"reviewed","updatedAt":"2026-08-10","title":"Amazon AI Capex in 2026: $200B Spending Guide","description":"Track Amazon's approximately $200 billion 2026 capex plan, the AWS and AI investment signals behind it, its companywide scope, and the cash-flow implications.","keywords":["Amazon AI capex 2026","Amazon capital expenditure 2026","AWS data center spending","AMZN AI infrastructure investment"],"guidance":{"period":"Calendar 2026","display":"About $200B","lowBillions":200,"highBillions":200,"midpointBillions":200},"latestSignal":{"label":"Q2 2026 cash-flow signal","value":"$66.1B","note":"Year-over-year increase in trailing property-and-equipment purchases, net of incentives"},"summary":"Amazon expects approximately $200 billion of companywide capital expenditure in 2026. Management connects much of the increase to AI and AWS demand, but the plan also funds chips, robotics, fulfillment, transportation and satellite infrastructure. It is therefore the least AI-specific headline number in the four-company comparison.","accountingScope":"Amazon's guidance covers the whole company rather than an AWS-only or AI-only capital budget. Its cash-flow statement reports property-and-equipment purchases net of proceeds and incentives, while the guidance discussed by management is a broader planning figure. Analysts should not label the full $200 billion as GPU or data-center spending.","allocation":["AWS data centers, networking and compute capacity for conventional cloud workloads and AI services.","Custom silicon programs including Trainium and Graviton, alongside third-party accelerators.","Robotics, fulfillment, transportation, devices and low-earth-orbit satellite infrastructure outside AWS."],"investorRead":"Amazon shows the cash cost of the buildout most visibly. In its second-quarter update, the company said trailing free cash flow had moved to a $7.6 billion outflow, driven primarily by a $66.1 billion year-over-year increase in property-and-equipment purchases net of incentives. The investment case depends on AWS and AI revenue eventually producing a durable cash-flow surplus large enough to justify both the cloud investment and the non-AWS projects included in the same capex envelope.","watchItems":["Purchases of property and equipment and the resulting free-cash-flow trajectory.","AWS growth, backlog and disclosures about customer commitments tied to new capacity.","Trainium, Graviton and AI-service revenue signals that indicate utilization of the installed base.","Evidence separating AWS infrastructure spending from fulfillment, robotics and satellite investment."],"limitations":["The $200 billion plan is companywide and cannot be treated as pure AI or AWS capex.","Cash investment, incentives, leases and management's planning definition are not perfectly interchangeable.","Customer commitments can support utilization but do not guarantee the return ultimately earned on every project."],"sources":[{"label":"Amazon Q2 2026 results","href":"https://www.aboutamazon.com/news/company-news/amazon-earnings-q2-2026-report","publishedAt":"2026-07-30"},{"label":"Amazon Q4 2025 and full-year results","href":"https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-Fourth-Quarter-Results/","publishedAt":"2026-02-05"}]},{"slug":"alphabet","company":"Alphabet","ticker":"GOOGL","searchStatus":"reviewed","updatedAt":"2026-08-10","title":"Google AI Capex in 2026: $175B–$185B Guide","description":"Track Alphabet and Google's $175 billion to $185 billion 2026 capex plan, server and data-center mix, Cloud allocation, depreciation, and investor signals.","keywords":["Google AI capex 2026","Alphabet capital expenditure 2026","Google data center spending","GOOGL AI infrastructure investment"],"guidance":{"period":"Calendar 2026","display":"$175B–$185B","lowBillions":175,"highBillions":185,"midpointBillions":180},"latestSignal":{"label":"2025 capex baseline","value":"$91.4B","note":"About 60% servers and 40% data centers and networking"},"summary":"Alphabet expects $175 billion to $185 billion of capital expenditure in 2026, nearly doubling the $91.4 billion invested in 2025 at the midpoint. Management says the mix should remain similar to 2025: approximately 60% servers and 40% data centers and networking equipment. The spending supports Google Cloud, DeepMind, Search, advertising systems and other businesses.","accountingScope":"Alphabet defines capital investment primarily around technical infrastructure and office facilities. The timing of supplier payments can move reported quarterly capex, and the company does not isolate one audited AI-capex line. Management expects just over half of 2026 machine-learning compute to be allocated to Google Cloud, but that allocation is a compute measure rather than a direct capex split.","allocation":["Servers and machine-learning compute for Google Cloud, DeepMind and first-party services.","Long-duration data centers and networking equipment needed to deliver that compute.","Infrastructure supporting Search, YouTube, advertising systems and selected Other Bets alongside external Cloud demand."],"investorRead":"Alphabet's disclosed 60/40 infrastructure mix creates a useful bridge between near-term server deployment and long-lived construction. Servers may translate into capacity and depreciation sooner, while buildings and networks create a longer placement-in-service schedule. Investors should compare Cloud growth and backlog with the accelerating depreciation and data-center operating costs management has already flagged.","watchItems":["Quarterly capex cadence relative to the $175 billion to $185 billion full-year range.","The server versus data-center and networking mix as infrastructure ramps.","Google Cloud growth, backlog and evidence that new capacity is relieving supply constraints.","Depreciation and data-center operating-cost growth across Alphabet's income statement."],"limitations":["Alphabet does not publish a clean AI-only capital-expenditure measure.","The disclosed compute allocation is not the same as a dollar capex allocation.","Payment timing and the placement-in-service schedule can make quarterly capex and depreciation move differently."],"sources":[{"label":"Alphabet Q4 2025 earnings call","href":"https://abc.xyz/investor/events/event-details/default.aspx","publishedAt":"2026-02-04"},{"label":"Alphabet investor FAQ: capital expenditures","href":"https://abc.xyz/investor/faqs-and-general-information/default.aspx","publishedAt":"2026-02-04"}]},{"slug":"meta","company":"Meta","ticker":"META","searchStatus":"reviewed","updatedAt":"2026-08-10","title":"Meta AI Capex in 2026: $130B–$145B Guide","description":"Track Meta's updated $130 billion to $145 billion 2026 capex outlook, finance leases, quarterly spending, advertising monetization, and investor signals.","keywords":["Meta AI capex 2026","Meta capital expenditure 2026","Meta data center spending","META AI infrastructure investment"],"guidance":{"period":"Calendar 2026","display":"$130B–$145B","lowBillions":130,"highBillions":145,"midpointBillions":137.5},"latestSignal":{"label":"Q2 2026 capex","value":"$31.08B","note":"Includes principal payments on finance leases"},"summary":"Meta narrowed its 2026 capital-expenditure outlook to $130 billion to $145 billion on July 29, raising the lower end from its earlier $125 billion to $145 billion range. The spending includes principal payments on finance leases and supports data centers, servers, networking, recommendation systems, generative AI and research infrastructure.","accountingScope":"Meta explicitly includes principal payments on finance leases in its capex measure. That convention is important when comparing the company with peers that describe additions, cash purchases and leases differently. Meta also monetizes much of its infrastructure indirectly through advertising engagement and efficiency rather than by selling every unit of compute to an external customer.","allocation":["Servers, accelerators and networking for recommendation, ranking and generative-AI workloads.","Data-center construction and finance-lease commitments needed to expand the installed compute base.","Infrastructure that supports the advertising business, consumer products, enterprise initiatives and long-horizon AI research."],"investorRead":"Meta's return bridge is different from a cloud provider's. New infrastructure can improve recommendations, ad conversion and product engagement without producing a separately reported AI revenue line. The analytical challenge is to compare those benefits with the rapid growth in depreciation, operating expenses, lease commitments and free-cash-flow demands created by the buildout.","watchItems":["Quarterly capex against the narrowed $130 billion to $145 billion outlook.","Finance-lease principal and future data-center commitments alongside cash purchases.","Advertising growth, recommendation improvements and any direct enterprise-AI monetization signals.","Depreciation, infrastructure operating costs and free cash flow as new capacity enters service."],"limitations":["The capex outlook includes infrastructure used across Meta's core business and research, not only generative AI.","Finance-lease treatment reduces comparability with peer headline numbers.","Indirect advertising benefits are harder to attribute to a specific infrastructure cohort than contracted cloud revenue."],"sources":[{"label":"Meta Q2 2026 results","href":"https://investor.atmeta.com/investor-news/press-release-details/2026/Meta-Reports-Second-Quarter-2026-Results/default.aspx","publishedAt":"2026-07-29"},{"label":"Meta Q2 2026 earnings event","href":"https://investor.atmeta.com/investor-events/default.aspx","publishedAt":"2026-07-29"}]}]}