Artificial intelligence is reshaping the enterprise software market
Artificial intelligence is changing the game in the corporate world, and we are not just talking about cute chatbots or virtual assistants. What is happening right now goes much deeper, and it is shaking the foundations of one of the most lucrative markets in tech: enterprise software.
The logic that dominated this sector for decades was pretty straightforward: more employees at a company meant more software licenses sold. Every person who joined an organization needed access to tools, platforms, and systems, and that translated directly into revenue for the big tech companies. It was a predictable, scalable, and extremely profitable model for whoever was selling those solutions.
But what happens when AI starts replacing some of those people? Fewer humans working could mean fewer paid licenses, which in turn would put the revenue of giants like Microsoft itself at risk. It is a scenario that seems simple on the surface, but it carries enormous implications for the software market as a whole, affecting everything from startups to the most established corporations on the planet.
That is exactly the fear keeping a lot of people in the industry up at night. And it is not unfounded. With the rise of agentic AI, market analysts have already started raising the question of how traditional licensing models will survive in an environment where autonomous AI agents take over tasks that used to be exclusively handled by human employees.
The good news, at least for Microsoft, is that a response to this problem seems to have emerged, and it comes from two heavyweights within the company. CEO Satya Nadella and Executive Vice President Rajesh Jha appear to be aligned on a vision that could completely flip the script. The central idea revolves around Microsoft Foundry and a new model where AI agents themselves become corporate users, each with their own identity, their own environment, and of course, their own license.
Sound controversial? Sure, but it also makes a lot of sense once you understand the logic behind it all. 👇
What is Microsoft Foundry and why it matters so much
Microsoft Foundry is, at its core, the infrastructure layer Microsoft designed to let companies create, manage, and deploy artificial intelligence agents within their corporate ecosystem. Think of it as a kind of AI agent factory, where organizations can configure and scale autonomous systems capable of executing complex tasks, making data-driven decisions, and interacting with other systems without needing constant human intervention. This is not a tool for building an assistant that answers emails. It is infrastructure for agents that operate as real digital workers, with context, memory, and the ability to take action.
The big announcement this week is a feature called Hosted Agents in Foundry Agent Service, which is already available in public preview. In practice, this feature gives each AI agent its own dedicated cloud computing environment, a kind of corporate sandbox with its own storage, its own identity, and its own set of permissions. To make it easier to picture, imagine every AI assistant getting its own locked office inside a high-security building, complete with its own badge and its own filing cabinet.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Nadella himself made a point of highlighting the concept when he announced the update on his X account. The key takeaway from the message was direct: every agent is going to need its own computer. And with the new Hosted Agents in Foundry, each agent gets its own dedicated enterprise-grade sandbox, with durable state, built-in identity and governance, plus support for any framework or harness.
The Foundry proposition is directly tied to Nadella’s vision for the future of enterprise productivity. Instead of simply offering tools that humans use to work better, Microsoft is betting on a model where AI agents themselves become active participants in the organizational structure. That means each agent can have a profile, permissions, access levels, and even a digital identity within the organization, just like any human employee would. It is a paradigm shift that goes well beyond a simple software update.
What makes Foundry especially relevant is that it is not just a technical bet but also a strategic response to the revenue problem we mentioned earlier. If AI agents start operating as users within corporate systems, they also need software licenses to access those platforms. With that, Microsoft turns a threat to its business model into a new growth vector, expanding the market instead of watching it shrink as human headcount declines.
Rajesh Jha’s argument: fewer humans, but more licenses
Rajesh Jha, Executive Vice President of Experiences and Devices at Microsoft, has been making an argument that directly tackles the fear dominating the corporate software sector. Earlier this month, Jha explained that if AI agents are going to operate inside enterprise software systems, they will need their own logins, their own inboxes, and their own digital identities. In other words, they will function as users.
And if they function as users, they may need to be licensed as users.
Jha used a well-known industry term when referring to this idea: seat opportunities. Every corporate agent with an identity and access represents a new seat — meaning a new paid license within the platform.
To make the concept even more concrete, Jha gave a pretty straightforward example. Imagine a company with 50 human employees paying for 50 software licenses. Now imagine that company automates 40 positions using AI agents and keeps only 10 human workers. Even in that scenario, the company would still be paying for 50 total licenses: 10 for the humans and 40 for the AI agents. The result? A company that downsizes to save money might end up spending the same amount, or even more, on software to keep the digital agents that replaced its workers.
This reasoning is what connects Jha’s vision with the Foundry update announced by Nadella. If every agent needs its own computer, its own identity, and its own secure environment, then the licensing ecosystem does not shrink. It simply changes in nature. Instead of licensing only people, platforms start licensing autonomous agents as well. It is a complete reconfiguration of what it means to be an enterprise software user.
AI agents as corporate users: the turning point in the licensing model
Here is the part that raises the most eyebrows, but also reveals the strategic brilliance behind this move. The idea that AI agents can be treated as corporate users, with their own identities and associated licenses, represents a complete reinterpretation of what a software user actually is. Historically, this concept was always tied to a real person — someone who logs in, performs tasks, and consumes platform resources. But with the arrival of autonomous agents capable of doing exactly that, the line between a human user and a digital user is starting to dissolve in a very tangible way.
In practice, what Microsoft is proposing with Foundry is that every agent deployed within an organization gets its own digital workspace. That includes access to tools like Microsoft 365, permissions within Azure, integrations with internal systems, and of course, a billing model tied to that usage. Instead of charging per human seat, the company begins charging for computational capacity, number of active agents, or volume of tasks executed. It is a transition from a per-head model to a per-utility model, and that completely changes the financial math for the entire sector.
For companies adopting this model, the impact on enterprise productivity could be significant. An AI agent does not need vacations, does not have limited working hours, and can operate across multiple processes simultaneously. If a company used to need ten licenses for ten financial analysts, tomorrow it might have a blended team of humans and agents, where each agent also has its own license but delivers processing power no human could match on their own. The cost per task drops, execution speed goes up, and Microsoft’s business model holds steady or even grows in this new landscape.
The real impact on software licenses and the corporate market
The question of software licenses in the context of corporate AI is one of the hottest debates in the tech industry right now. For decades, the enterprise software market was built on a pretty stable premise: headcount growth meant revenue growth for platforms. SAP, Salesforce, Oracle, Microsoft — all of these giants calibrated their pricing models around that logic. Now, with intelligent automation advancing at a rapid pace, that premise is being questioned in boardrooms around the world.
What Microsoft is doing with Foundry is essentially creating a new customer category: the corporate agent. And by doing so, it not only protects its existing revenue base but also opens up room to expand the licensing market in ways that did not exist before. An agent that accesses SharePoint data, processes information in Azure, generates reports in Excel, and communicates via Teams is, in practice, a heavy user of the Microsoft platform — and it makes perfect sense for that usage to be monetized. The question is how this will be structured and priced over the next few years, something the market is still figuring out.
Beyond that, this move is likely to pressure other market players into rethinking their own models. If Microsoft manages to validate the idea of licenses for AI agents in the corporate environment, competitors like Google, Salesforce, and SAP will probably follow similar paths. That creates a new market standard where artificial intelligence adoption does not reduce platform revenue but transforms it and potentially amplifies it. For companies on the customer side, this means a whole new set of decisions around budgeting, governance, and automation strategy that will need to be made much more carefully than simply buying a basic email license.
Security, governance, and identity: the pillars of the new model
One detail that should not be overlooked in this story is the emphasis Microsoft is placing on security and governance within Foundry. When Nadella says every agent will have its own enterprise-grade sandbox with built-in identity and governance mechanisms, he is signaling that the company understands the biggest obstacle to mass adoption of AI agents in the corporate world is not the technology itself — it is trust.
Companies need to be confident that an autonomous agent operating within their systems will not access data it should not, make decisions outside its defined scope, or create security vulnerabilities. By providing an isolated and controlled environment for each agent, with durable state and granular permissions, Foundry aims to address exactly that concern. It is a layer of trust infrastructure that could prove decisive in convincing CIOs and CTOs to get on board with this new model.
This focus on governance also reinforces the idea that AI agents are not just tools but operational entities that need to be managed with the same rigor applied to human employees. Access policies, usage audits, permission limits — all of that needs to exist for the model to work in a sustainable and secure way. And Microsoft seems to be building exactly that foundation with the Foundry update.
What this means for the future of corporate work
What is taking shape here is a deep reconfiguration of how enterprise software will be purchased, used, and billed in the coming years, and Microsoft Foundry appears to be well positioned to lead that transition.
Corporate artificial intelligence is no longer just a promise of efficiency. It is becoming a new layer of organizational infrastructure, with its own users, its own access levels, and now, its own licenses. The alignment between Satya Nadella and Rajesh Jha on this topic shows that Microsoft is not just reacting to market fears but trying to set the rules of the game for the next era of enterprise software.
For tech professionals, managers, and decision-makers, the message is clear: the per-seat licensing model is not going away — it is going to expand to accommodate a new class of digital users. Organizations that understand this early and start planning their automation strategies with this reality in mind will be in a position to capture productivity gains without being caught off guard by the associated costs.
Microsoft’s play with Foundry might seem bold today, but within a few years it could very well become the standard for how tech companies structure their relationship with corporate clients in an era dominated by autonomous agents. 🤖
