While the company has demonstrably succeeded in deploying a range of beneficial generative AI (genAI) tools and services, their adoption rate within enterprises continues to be surprisingly modest.
A common belief has emerged that we are currently experiencing an AI bubble. Those skeptical of AI often reference a McKinsey report, which indicated that “almost eight out of ten companies utilize gen AI — yet an equal number report no substantial impact on their bottom line.” Furthermore, they frequently mention an MIT study, The GenAI Divide: State of AI in Business 2025, which asserted that 95% of generative AI (genAI) pilot projects in businesses are unsuccessful.
However, an increasing body of evidence suggests that businesses are, in fact, realizing substantial benefits from AI, indicating that the notion of an AI bubble might be unfounded.
A December report by venture capital firm Menlo Ventures, titled 2025: The State of Generative AI in the Enterprise, revealed that “despite concerns about excessive investment, AI is being deployed across businesses at an unparalleled rate in the history of modern software…. Our most recent market intelligence highlights widespread adoption, tangible revenue, and significant productivity enhancements, pointing to a boom rather than a bubble.”
In which sectors is AI experiencing rapid growth?
This comprehensive report delivers much encouraging information for proponents of AI. It disclosed that AI expenditure reached $13.8 billion in 2025, representing over a sixfold increase from the $2.3 billion invested in 2023. The report characterizes this as “a distinct indication that organizations are moving beyond experimental phases to active implementation, integrating AI into their fundamental business approaches…. [Seventy-two percent] of key decision-makers foresee widespread adoption of generative AI tools in the foreseeable future.”
The report asserts that the emerging boom in genAI is evident not merely in the financial outlay, but also in the deployment patterns. Initial growth was primarily driven by conversational AI platforms such as Microsoft’s Copilot and OpenAI’s ChatGPT, which powers Copilot. However, the report now indicates that enterprise AI is “progressively being adopted across various departmental applications, including coding, sales, customer service, human resources, and specialized sectors ranging from healthcare and legal services to the creative industries.”
Crucially, businesses have fundamentally altered their approach to AI development and implementation. Instead of creating proprietary tools in-house, they are procuring them from specialized AI providers. The report states that in 2024, 53% of AI solutions were purchased externally rather than developed internally, a figure that soared to 76% by 2025. The report highlights that “off-the-shelf AI solutions are reaching operational status more rapidly and delivering instant value,” contrasting with internal projects that often require longer development cycles and do not always yield the anticipated returns.
What is Microsoft’s stake in this growth?
What implications does this hold for Microsoft? Evidence suggests that the company may not gain as significantly from this boom compared to other AI enterprises. The financial news outlet The Motley Fool cautions, “The financial figures supporting Microsoft’s AI expansion appear inconsistent.” It observes that while Microsoft’s AI division seems to be performing robustly, with $625 billion in service orders for its Azure cloud platform, a substantial $281 billion of this amount originates from OpenAI, a company reportedly struggling against rivals and potentially facing fiscal difficulties.
Microsoft often highlights the achievements of Microsoft 365 Copilot, its generative AI feature integrated into the Microsoft 365 business productivity suite. At the conclusion of the previous fiscal quarter, the company reported 15 million paid subscriptions for Microsoft 365 Copilot. While this figure seems considerable, The Motley Fool notes that Microsoft 365 itself accounts for 450 million paid users. Thus, “despite Microsoft’s substantial AI investments, Copilot’s adoption rate barely surpasses 3%. Business clients seem exceptionally hesitant to subscribe to Copilot.”
Likewise, the article indicates that GitHub Copilot, an AI-powered coding assistant, has only 4.7 million paid subscribers, despite GitHub itself having around 150 million users — a usage rate also just above 3%.
Microsoft’s counter-strategy
Acknowledging the necessity to improve these statistics, Microsoft has unveiled several new Copilot offerings designed for corporate clients. Copilot Cowork for Microsoft 365 transcends mere question-answering or basic interaction with Microsoft 365 applications; it executes tasks according to user needs. For instance, in preparation for a client meeting, it can “extract pertinent information from emails, meetings, and documents, allocate preparation time on the calendar, and subsequently generate a comprehensive package of materials, including a briefing document, supplementary analysis, and a presentation deck ready for the client.”
Additionally, on Monday, the company launched Microsoft Agent 365, a system for overseeing enterprise AI agents, which automatically establishes each agent’s identity, access rights, operational scope, and risk profile. Microsoft is also deploying a new enterprise package, 365 E7: The Frontier Suite, integrating Microsoft 365 with both Microsoft 365 Copilot and Agent 365.
While these initiatives are commendable, their ability to significantly increase Copilot’s current low adoption rate remains questionable. It is probable that Copilot will eventually be incorporated as a free or nearly free component of Microsoft 365 enterprise subscriptions, mirroring Google’s strategy with Gemini and Google Workspace. Such a move would undoubtedly boost Copilot’s usage, but it would not generate supplementary income.
Therefore, does Microsoft face an era of significant growth or an eventual decline? Much hinges on the performance of Azure and whether OpenAI sustains its reliance on the Azure services it currently utilizes.
The most significant unknown, however, is the success of Microsoft’s investment in what it terms “Humanist Superintelligence.” This concept involves a collection of AI-driven technologies, each designed to address a critical issue and improve human well-being. Should this endeavor prove successful, the company could experience truly prosperous times. Conversely, failure could ultimately lead to Microsoft losing its position as the global leader in AI innovation.
