Microsoft reports surge in enterprise AI adoption and Copilot usage

From novelty to core operations

Microsoft's latest fiscal first-quarter earnings call revealed a significant shift in how large corporations are using artificial intelligence. CEO Satya Nadella noted that companies are moving away from using AI for isolated productivity tasks and are instead applying it to solve operational problems. This transition marks a move from "shiny object" experimentation to practical, functional integration within large-scale business structures.

The company highlighted several major global brands as examples of this trend. Unilever is currently using AI to automate its order intake and manage disputes regarding trade promotions. These are core logistical functions that involve high volumes of data and repetitive decision-making processes that were previously handled manually.

Other large-scale applications include Rolls-Royce, which uses AI to manage maintenance planning and supply chains for jet engines containing roughly 30,000 parts. Additionally, the Austrian lender Raiffeisen Bank International has deployed AI across more than 20 different business processes. These examples demonstrate that enterprise AI is being used to manage complex inventory and credit-related workflows rather than just assisting with individual writing tasks.

Surging Copilot adoption rates

The data regarding Microsoft Copilot usage shows a rapid acceleration in adoption across the corporate world. Daily users of the tool rose nearly 50% from the previous quarter, which represents a more than 5x increase compared to the same period last year. This growth suggests that the initial barrier to entry for AI tools in the workplace is falling as more departments find viable use cases.

The scale of this adoption is particularly evident among very large organizations. The number of customers with more than 50,000 Copilot seats has grown more than 8x over the past year. This indicates that massive enterprises are committing to wide-scale deployment rather than just testing the technology in small, isolated teams or departments.

Automated tasks have also seen a significant jump, nearly doubling since July. This trend is supported by the fact that agents now open more than a third of all pull requests on GitHub. Furthermore, the number of agent-generated pull requests has increased 6x in just six months, showing how deeply automated systems are becoming part of the software development lifecycle.

New consumption-based billing

To accommodate these more complex uses, Microsoft is changing how it charges for its AI services. While standard subscriptions will continue to provide everyday AI tools, the company is adding consumption-based billing for agents that handle more complex tasks. This means that while a user might pay a flat fee for basic access, the company will charge extra based on the actual work the AI agents perform.

This model reflects a growing distinction between general-purpose AI and specialized agents. IT departments typically fund the standard subscriptions that provide everyday AI for the workforce. However, individual business units may now have the option to pay specifically for agents that perform higher-value, more intensive work that requires more computing resources or complex logic.

The shift toward this model is already showing early results, with the number of Copilot customers consuming credits through usage-based billing growing more than 75% in the last quarter. CFO Amy Hood noted that consumption pricing outside of software development will likely move more slowly because it requires changes in sales practices and significant customer education to understand the new structure.

Infrastructure and security

As enterprises move toward more complex automation, the underlying infrastructure and security of these systems have become a priority. Microsoft reported that customers have used Microsoft Purview to audit more than 7.6 billion agent activities. This level of auditing is necessary for companies to ensure that AI systems are operating securely and following internal compliance rules.

The company is also standardizing how these agents are governed on Windows systems. Microsoft Execution Containers now govern agents, a move that is being adopted by major AI developers. Specifically, OpenAI Codex, Anthropic Claude Code, and Meta Muse are all adopting these containers to manage how their agents run in enterprise environments.

The scale of the agent ecosystem is expanding rapidly as well. The Agent 365 registry currently lists more than 50 million agents, which is an increase of more than 30% from the previous quarter. Meta is also beginning to deploy these execution containers across its enterprise offerings to ensure consistency and reliability for its users.

Financial performance and growth

The growth in AI adoption is contributing to a strong financial performance for Microsoft. The company reported total revenue of $92.3 billion for the quarter, which is a 19% increase. This growth was driven largely by the cloud sector, with Microsoft Cloud revenue reaching $63.9 billion, a 30% increase over the previous period.

Azure, Microsoft's primary cloud computing platform, saw an even larger jump with revenue rising 48%. These numbers suggest that the infrastructure powering AI is seeing massive demand as more companies transition their data and workflows to the cloud. This growth is a key indicator of the underlying demand for the computing power required to run large language models and autonomous agents.

On the consumer side, Microsoft 365 consumer cloud revenue rose 17%. However, this growth was slower than the previous quarter as the company began lapping an earlier price increase. While the consumer market remains steady, the primary engine of growth appears to be the enterprise sector where AI integration is accelerating at a faster pace than in the general consumer market.