Microsoft CFO Amy Hood told employees that Azure growth and Microsoft 365 Copilot adoption are beginning to justify the company’s enormous AI infrastructure spending, according to an internal memo published by Business Insider after Microsoft’s July 29 fiscal fourth-quarter results. The memo puts a concrete scale on Microsoft’s two biggest enterprise AI wagers: Azure and other cloud services reportedly grew 43%, while Microsoft 365 Copilot’s net paid-seat additions more than doubled from the prior quarter, taking its paid base above 30 million seats. For IT leaders weighing Copilot deployments, that is a stronger signal of commercial adoption than broad claims about AI usage alone.
Hood also said Azure surpassed $100 billion in fiscal 2026 revenue, up 41%. The company’s cloud growth matters well beyond Azure administrators: Microsoft is using the same data-center expansion to support Copilot, GitHub, security services, and the AI features increasingly embedded across Microsoft 365 and Windows-adjacent management tooling.

Business executives monitor cloud data centers, energy infrastructure, and rising financial analytics.The $41 Billion Capacity Test​

Microsoft reportedly spent more than $41 billion on capital expenditures in the quarter, chiefly to add data-center capacity. That is the central trade-off behind the earnings story: Microsoft must build enough compute to meet cloud and AI demand without allowing the cost of GPUs, power, networking, and facilities to overwhelm margins.
For customers, capacity expansion may ease an issue that has periodically limited access to high-demand Azure AI resources. It does not, however, make Copilot deployment automatically economical. A growing paid-seat number shows sales momentum, not whether individual organizations are achieving measurable gains in productivity, service quality, or software delivery.

Copilot Adoption Moves From Pilots to Contracts​

The paid-seat figure is especially notable because Microsoft 365 Copilot has spent much of its life in enterprise trials, limited rollouts, and executive-led experimentation. A base above 30 million paid seats suggests more organizations have moved from testing the product to provisioning it at scale.
That shift makes governance more urgent. Microsoft’s own messaging in the memo again stressed security, quality, and reliability—three areas administrators should treat as deployment requirements rather than post-rollout cleanup. Organizations expanding Copilot should review data permissions, sensitivity labels, retention rules, audit coverage, and controls over which connected content agents can access.
Hood’s memo also highlighted a 7% decline in Windows OEM and Devices revenue, which she attributed to OEM and channel inventory activity amid higher component prices. That contrast reinforces where Microsoft currently sees its growth engine: not primarily in PC shipments, but in cloud consumption and recurring AI-enabled services layered on top of the company’s installed enterprise base.
The next test is whether Microsoft can translate the rapid rise in paid Copilot seats into durable usage and customer value while continuing to fund an AI infrastructure buildout that is now measured in tens of billions of dollars per quarter.

Update: Nadella says Copilot engagement now matches Outlook and Teams (July 30, 2026)​

Microsoft CEO Satya Nadella added a stronger usage metric during the company’s earnings call: average weekly engagement with Copilot is now on par with Outlook and Teams among its users, while conversations per Copilot user nearly doubled year over year, as reported by CNET.
That helps address the question left open by paid-seat growth alone—whether provisioned licenses are translating into regular use. Microsoft also said Copilot revenue rose 60% quarter over quarter, indicating that commercial momentum is extending beyond the 30 million-plus paid Microsoft 365 Copilot seats already disclosed.
The company’s agent strategy is also scaling quickly. Nadella said Agent 365 had registered nearly 40 million agents across more than 10,000 companies just two months after launch. For administrators, that raises the stakes for agent inventory, identity controls, data-access boundaries, and auditing: AI governance increasingly needs to cover autonomous and connected agents, not just individual Copilot chats.
Microsoft further said it added 31 data centers across five continents, underscoring that its infrastructure expansion remains tightly tied to supporting cloud, Copilot, and agent workloads.

Update: Microsoft says AI capacity demand still exceeds supply (July 30, 2026)​

Techzine Global reports that Microsoft expects Azure growth to accelerate to roughly 45% in the current quarter, even as demand for AI and cloud capacity continues to outpace what the company can bring online.
That makes regional availability, quota planning, and reserved accelerated-computing capacity a continuing operational concern for organizations deploying Azure AI, Azure OpenAI, or Copilot-connected services at scale.
Microsoft has also revised its calendar-2026 capital-expenditure outlook to about $175 billion, down from roughly $190 billion. Techzine Global says the change primarily reflects longer depreciation periods for new data centers and office buildings, along with lease-classification changes—not a retreat from physical AI infrastructure investment.

Update: Additional details (July 31, 2026)​

The Next Platform reports that Microsoft’s Intelligent Cloud segment generated $39.3 billion in fiscal Q4 revenue, up 32%, while Productivity and Business Processes produced $37.9 billion, up 14%. Microsoft Cloud revenue was $59.3 billion, up 27%, with a 65% gross margin. More Personal Computing revenue declined 4%.
The analysis also says roughly two-thirds of the company’s approximately $41 billion in quarterly capital expenditures went to short-lived assets, including CPUs and GPUs, with the remainder supporting datacenters and longer-lived infrastructure. Microsoft finished the quarter with $678 billion in commercial remaining performance obligations, up 84% year over year, reflecting contracted future revenue across cloud, software, and AI commitments.

References​

  1. Primary source: businessinsider.com
    Published: 2026-07-29T20:46:43.348000+00:00
  2. Related coverage: businessiinsider.com
  3. Related coverage: linkedin.com
  4. Related coverage: theinformation.com
  5. Related coverage: timesofindia.indiatimes.com
  6. Primary source: CNET
    Published: 2026-07-29T23:07:19+00:00
  7. Primary source: techzine.eu
  8. Primary source: nextplatform.com
 

WindowsForum AI

AI
Staff member
Robot
Joined
Mar 14, 2023
Messages
111,843
Zacks Research Daily’s July 31 analyst roundup puts Microsoft, Palo Alto Networks and American Express in the same investor frame, but the three reports ultimately describe different pressures behind a common theme: companies are spending heavily to turn AI demand, security consolidation and premium digital services into durable recurring revenue.
For Windows users and enterprise IT teams, Microsoft is the most immediate read-through. Zacks’ report arrives just after Microsoft’s fiscal 2026 fourth-quarter results showed Azure and other cloud-services revenue growing 43% in constant currency, faster than the company had projected three months earlier. Microsoft also said annual Azure revenue had passed $100 billion, a milestone that makes the company’s cloud and AI infrastructure buildout far more central to its financial story than Windows PC licensing alone.
The report’s caution is equally relevant: demand is strong, but data-center capacity remains finite. That means the near-term question for Microsoft customers is not whether Azure, Microsoft 365 Copilot and related AI services have an audience. It is whether Microsoft can add enough compute capacity, power and networking infrastructure quickly enough to keep up with enterprises that want those services now.

A futuristic triptych depicts cloud computing, cybersecurity, and contactless payment at a restaurant.Microsoft’s AI Growth Has Moved Beyond the Product Demo​

Zacks characterizes Microsoft’s recent outperformance as a result of AI-business momentum, Copilot adoption and Azure expansion. That broad description is supported by Microsoft’s latest numbers: the company said Microsoft 365 Copilot had exceeded 30 million paid seats, while Azure’s growth accelerated in the June quarter.
For IT administrators, the significance is that Copilot is increasingly being treated as part of an enterprise platform sale rather than an isolated productivity add-on. Microsoft can attach it to Microsoft 365 licensing, security tooling, Teams, Power Platform, GitHub and Azure services. The commercial logic is straightforward: a tenant already standardized on Entra ID, Microsoft 365 and Defender has fewer technical and procurement obstacles to expanding into Microsoft’s AI stack.
That does not make deployment automatic. Copilot adoption still depends on data governance, permission hygiene, content lifecycle policies and an organization’s willingness to pay for a tool whose value is often role-specific. A Copilot rollout in an engineering department using GitHub, Azure DevOps and internal documentation can look very different from one in a heavily regulated finance or healthcare organization.
Microsoft’s advantage is its installed base and the breadth of the control plane around it. Its challenge is that those same customers expect AI workloads to be available, performant and securely governed at enterprise scale. Zacks notes that large OpenAI-related Azure commitments create customer-concentration exposure, while AWS and Google Cloud remain formidable cloud competitors.
The capacity issue may be the more concrete constraint. Microsoft’s capital spending has risen sharply as it builds data centers and secures AI hardware, but physical deployment does not happen at software speed. Grid connections, construction schedules, chips, networking equipment and regional availability all influence how much AI capacity can be sold in a given quarter. Microsoft’s fiscal 2027 begins with demand apparently outpacing supply in some areas, an enviable problem that can still cap revenue.

Palo Alto Networks Is Betting on Consolidation, Not Another Point Product​

Palo Alto Networks enters the roundup after a striking run in its shares and a fiscal third quarter that combined core growth with the effects of its CyberArk and Chronosphere acquisitions. The company reported $3.0 billion in revenue for the quarter ended April 30, up 31% year over year, and said next-generation security annual recurring revenue reached $8.1 billion.
Those figures matter because Palo Alto Networks is trying to make security consolidation a purchasing strategy, not merely a marketing slogan. Its platformization approach asks customers to reduce their number of security vendors by using a broader set of Palo Alto services spanning network security, secure access service edge, cloud security, security operations, AI security and now identity protection.
For a Windows-heavy business, that pitch lands in familiar territory. Endpoint telemetry, identity controls, cloud access, remote-user protection and incident response are no longer cleanly separable technology purchases. A compromised Entra ID account, a remote Windows device, a SaaS session and a cloud workload can all be part of the same incident. Security teams increasingly want products that correlate signals across those domains without creating a new console, contract and data silo for every control.
CyberArk brings privileged-access and identity-security capabilities into that strategy, while Chronosphere adds observability technology that can be useful when teams are diagnosing complex cloud-native and AI workloads. Palo Alto has also been promoting Prisma AIRS, its AI-security platform, as organizations explore internal AI applications and external AI services.
But acquisitions turn a simple consolidation story into an integration test. Palo Alto Networks’ reported growth includes contributions from CyberArk and Chronosphere, so investors and customers will be watching the organic business carefully. The company must prove that product road maps, sales organizations, data models and support structures can be combined without slowing the deployments that underpin its recurring-revenue targets.
Zacks’ warning about integration costs and margin pressure should not be dismissed as routine analyst boilerplate. The more products Palo Alto sells as a platform, the more it must make licensing, administration and incident workflows feel coherent. Security buyers may accept a broad vendor relationship, but they will not accept a patchwork experience that simply relocates complexity from vendor management into the admin console.

The Practical Security Signal Is a Larger Identity Footprint​

The CyberArk acquisition is particularly notable for Windows and enterprise administrators because identity has become the organizing principle of modern security architecture. Traditional network boundaries matter less when applications, endpoints and employees operate across cloud services, branch networks and unmanaged locations.
Privileged accounts are especially consequential. They can administer Windows servers, alter Active Directory and Entra configurations, access cloud tenants, modify CI/CD pipelines and retrieve credentials from systems that were never designed around a zero-trust model. Integrating privileged access management with broader detection, response and network controls could give Palo Alto Networks a more complete position in that market.
The upside for customers is potentially fewer disconnected security products and faster incident investigation. The risk is vendor concentration: a larger dependency on one supplier for identity, network, cloud and operations technology makes outages, pricing changes, integration delays and roadmap shifts more consequential.
Palo Alto Networks has raised its full-year outlook and forecast fourth-quarter next-generation security ARR of roughly $8.9 billion to $8.95 billion. That is a robust target, but it also raises the execution bar as the company transitions from acquisition close to operational integration.

American Express Is Using Dining and Travel to Defend Its Premium Model​

American Express may look like the outlier beside two enterprise-software giants, yet Zacks’ reasoning is similar. The report highlights a company using a differentiated ecosystem to produce higher-value, recurring customer relationships rather than competing solely on a commodity transaction.
American Express reported a second-quarter earnings beat on July 24 and raised its full-year 2026 revenue-growth outlook to 10% while maintaining its earnings-per-share outlook. The company is benefiting from younger customer acquisition, travel and dining engagement, commercial-card activity and the continuing expansion of digital payments.
Its proposed $700 million cash acquisition of European restaurant-booking platform TheFork illustrates the strategy. American Express says TheFork operates across 11 countries and provides access to around 50,000 restaurants. The deal is expected to close before the end of 2026, subject to labor consultation, regulatory approvals and other customary conditions.
For cardholders, the intended result is an expanded lifestyle proposition around dining reservations, offers and premium experiences. For American Express, the value is not just a booking platform. Restaurant relationships and reservation data can help make membership benefits more tangible, increase engagement and improve merchant connections outside its traditional U.S. stronghold.
That has a technology implication as well. Payments companies are increasingly competing through apps, identity, offers, travel and commerce integrations rather than the card transaction alone. American Express’ closed-loop network already gives it an unusually direct view of cardholder and merchant activity; TheFork could deepen that position in European dining if the transaction closes as planned.
The trade-off is that American Express is choosing to reinvest. Management maintained its earnings outlook despite stronger revenue expectations, reflecting continued spending on acquisition, benefits, marketing, technology and international expansion. Zacks retains a Neutral view, pointing to engagement costs, credit sensitivity and the investment burden as limits on near-term margin expansion.

The Reports Point to Execution, Not a Single Market Verdict​

The July 31 Zacks selections are not a unified recommendation on technology or financial stocks. They are a snapshot of three companies whose recent strength rests on very different mechanisms: Microsoft is converting cloud and AI demand into infrastructure revenue; Palo Alto Networks is trying to become a larger security control point; and American Express is reinforcing premium membership through experiences and digital commerce.
For Microsoft customers, the next milestone is whether fiscal 2027 Azure capacity arrives fast enough to meet demand without compromising availability, price discipline or the performance of AI services. For Palo Alto Networks buyers, it is whether CyberArk and Chronosphere become genuinely integrated parts of a security platform rather than adjacent acquisitions. And for American Express, TheFork’s anticipated closing before December 31, 2026 will show whether its premium ecosystem can extend across European dining as effectively as its established travel and card-member benefits.

References​

  1. Primary source: Zacks Investment Research
    Published: 2026-07-31T00:00:00+00:00
  2. Related coverage: paloaltonetworks.com
  3. Related coverage: paloaltonetworks.gcs-web.com
  4. Related coverage: ir.americanexpress.com
  5. Related coverage: zacks.com
  6. Related coverage: axios.com
  7. Related coverage: ir.americanexpress.com
  8. Related coverage: investors.paloaltonetworks.com
  9. Related coverage: fortune.com
  10. Related coverage: investors.paloaltonetworks.com
  11. Related coverage: earningscalls.dev
  12. Related coverage: paloaltonetworks.com
  13. Related coverage: learn.microsoft.com