Redesigning Organizations in the Age of AI

For a long time we have discussed the impact of AI on business in terms of efficiency. Faster customer service, shorter analysis cycles, easier access to enterprise knowledge, employees freed from repetitive tasks. These are all meaningful gains. Yet as generative AI and AI Agents mature, a more fundamental question comes into view: will AI only help us do our existing work better, or will it require us to rethink how the work itself is done?

This is exactly why the AI and education report MIT published in August 2026 deserves attention from the business world. The report was written with education in mind, but the approach it sets out opens a productive line of thinking for enterprises as well. MIT does not simply debate how AI should be used within existing teaching processes. It revisits the structure, meaning and value of education in a world where AI is advancing rapidly.

We can make a similar assessment for companies today.

Is adding AI to a process enough?

The Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training, formed within MIT, was initially charged with three topics: to assess how students and instructors use AI, to examine innovations in teaching and assessment methods, and to recommend policy on AI use.

Over time, however, the committee's work turned toward more fundamental questions. As MIT's leadership emphasized when presenting the report, the discussion extended to rethinking the structure, meaning and value of an MIT education in the age of AI.

This tells us something important. Major technological shifts call not only for the adoption of new tools but for a reassessment of existing structures.

At CBOT we encounter a similar pattern in enterprise AI projects. Many projects naturally begin by solving a specific problem inside an existing process. Understanding customer requests faster, simplifying access to enterprise knowledge, supporting employees in content production, or using time more efficiently in operational processes are some of them.

These are sound starting points. But when we want to use the full potential of generative AI, we need to ask the next question:

If we were designing this process for the first time today, knowing that people and AI would work together, would we still design it the same way?

This question moves AI transformation out of the category of a technology project and makes it part of organizational design.

The real subject is the division of labor between people and AI

Measuring the value of AI in an enterprise by automation rate alone is not the right approach. The real value appears when we help people reach information faster, decide with more context, and spend their time on higher quality work.

Consider customer service.

During a conversation, an agent may review a customer's past transactions, search for information across different systems, check procedures, and only then construct a resolution. In a setup supported by generative AI, the system can bring these pieces together and present them to the agent with the right context.

The core gain here is not only a shorter response time. The agent's time shifts from searching for information to understanding the customer, weighing alternatives and creating a better experience.

A similar shift is visible in finance, human resources, operations and sales teams.

A finance specialist focusing on evaluating results rather than hunting for information across dozens of pages, an HR team spending more time on employee experience instead of repetitive questions, or a sales team concentrating on understanding customer needs rather than gathering data all rest on the same design principle.

The aim is not to remove people from the process. It is to make the points where human contribution matters most more visible.

Workflows need to be rethought

Before generative AI, many enterprise processes were built on linear workflows in which people carried information between systems.

A request arrives. It is classified. It is routed to a team. Information is gathered. Rules are checked. A response is prepared. Another team joins the process when needed.

Generative AI and AI Agents make it possible to design this structure differently.

A system can understand the intent of a request, find the relevant enterprise knowledge, combine data from different sources, and offer the user or the employee options that fit the context. People remain at the center of the process where exceptions, judgment, prioritization and accountability are required.

The effect of this shift is not only operational speed. The structure of the workflow changes.

For this reason, building an enterprise generative AI strategy solely around the question "in which use cases can we apply AI?" may not be enough. The question "which processes can we redesign around the new capacities AI offers?" belongs on the agenda as well.

Three lessons for enterprises from MIT's approach

One of the striking aspects of the MIT report is that it does not treat technological transformation as a matter of tools alone. The report recommends making educational processes AI-aware, strengthening a structure that keeps people and community at the center, and establishing mechanisms for continuous assessment and improvement in the face of fast-moving technology.

These three approaches can be adapted to an enterprise AI strategy.

First, processes need to be made visible.

Before asking where a process sits on the organization chart, we should understand how it actually works. Which tasks consume people's time? Which systems do they move between to reach information? Which decisions repeat? Where does context get lost?

This analysis is an important starting point for seeing where AI can create value.

Second, the division of labor between people and AI needs to be designed.

Not every task needs to be carried out with AI. What matters is identifying where the speed, scale and processing capacity of AI combine best with human experience, judgment, communication and contextual knowledge.

The question for a successful generative AI project is therefore not "how much did we automate?" but "how do people and AI do this work better together?"

Third, a continuously learning structure needs to be established.

AI technologies are advancing quickly. The models, capabilities and methods in use today can change within a short period. This is why MIT recommends structures that support continuous assessment, experimentation and improvement rather than one-off solutions.

On the enterprise side, the equivalent is measuring and improving AI systems throughout their use, not only at the moment they go live.

Accuracy, user experience, business outcomes, cost, security and employee feedback should be monitored regularly within the same system.

AI transformation is an organizational subject

At CBOT we have been running AI projects since 2015 in high transaction volume organizations, from banking and finance to retail and aviation, from customer service to the public sector.

This experience shows us that AI transformation can begin with technology teams but cannot remain limited to them.

As generative AI and AI Agents spread, collaboration between operations, customer experience, human resources, finance, technology and business units becomes more important. Because we are no longer only choosing a technology. We are revisiting how work proceeds, how information flows and how decisions are supported.

This perspective matters especially for companies with large operations, dense enterprise knowledge, a high volume of customer interactions and interdependent processes.

What MIT does in its report is a similar shift in thinking. Rather than treating AI as a new tool added to the existing education model, it preserves the institution's core purposes and reassesses how those purposes should be achieved. MIT President Sally Kornbluth, noting that AI marks a turning point for education and research, also emphasizes that MIT's mission of developing people's capacity to explore, solve problems and create will remain central.

A similar approach is valuable for the business world.

The opportunity AI offers is not to leave people outside the organization. It is to strengthen their access to information, support decision quality, reduce the burden of repetitive work, and let teams focus on areas where they can create more value.

For this reason, one of the important questions of AI strategy in the period ahead will be, alongside "which model will we use?", this one:

If we were redesigning a company today where people and AI work together, how would we organize the work?

Asking this question properly can move AI out of the category of a technology added to an existing organization and make it a natural part of how the institution develops.