Microsoft restructures its divisions toward Agents and injects $45 billion into computing
We tracked in Microsoft’s books a record revenue for the tenth consecutive quarter, reaching $92.3 billion, a 19 percent year-over-year increase, at a time when the group completely reorganized its business reporting to place the “Agents and Infrastructure” unit at the top of its operating model. This growth was accompanied by a 32 percent jump in net income to $36.6 billion under generally accepted accounting principles, driven by the rise of the cloud segment, which alone generated $63.9 billion, a 30 percent annual growth.
Cloud revenue was split between the Office 365 platform, which contributed about $28 billion, and the Azure platform, which captured the larger share at $33.2 billion. Although Azure hosts traditional enterprise software workloads, artificial intelligence cemented its direct financial weight through a $10 billion flow of revenue from OpenAI alone into Microsoft’s books, alongside hosting Anthropic’s workloads. In contrast, the devices and consumer segment contracted by three percent to $15.5 billion, under a sharp decline in Windows licensing sales for devices of 22 percent.The bulk of corporate technology spending shifted from desktop computers to agent servers and cloud infrastructure.
This financial shift is accompanied by two indicators that are of negligible significance in the company’s books; the first is a 73 percent jump in remaining commercial performance commitments to $677 billion, which are binding future contracts representing reserved commitments from AI labs and commercial customers. To meet these massive obligations, the company’s quarterly capital expenditure rose to $45.2 billion, marking the highest level in its quarterly history.The jump in future performance commitments to $677 billion imposes a record capital race to build data centers.
For institutions and banks in the Gulf states, and tech companies in Egypt and the Levant, this landscape means that cloud and AI contracts are moving from exploratory trials to pre-reserved commitments. The acquisition of large shares of supercomputing platform capacity by major model labs, and the allocation of record budgets toward agent-infrastructure, require technology managers and digital-transformation leaders in the region to restructure procurement plans by reserving inferencing capacity through longer-term contracts to avoid cost volatility or availability bottlenecks. The restructuring of global business units also demands abandoning expansion in traditional device budgets and focusing local spending on enabling teams to manage automated agent environments and distributed computing platforms.
These figures show that AI revenues are no longer a deferred future bet, but have become an actual driver handling cash flows exceeding $90 billion quarterly. However, the continuation of this profitability remains conditional on the capital infrastructure’s ability, which consumes more than $45 billion every three months, to convert hundreds of billions of deferred commitments into sustainable productive value for companies.