? AI is quietly changing the traditional corporate hierarchy, flattening structures and reshaping job roles from the bottom up. Companies like Amazon, Moderna, and McKinsey are eliminating layers of middle management, merging departments, and deploying AI agents to automate routine work. As AI automates various tasks, even the C-suite is changing, with new AI leadership roles emerging and long-held power dynamics shifting.

The traditional corporate org chart, a neat triangle of power with executives at the top and junior workers at the base, is undergoing a quiet revolution, thanks to AI.

At Moderna, HR and tech now live under the same roof overseen by one Chief People and Digital Officer. At another one AI-first healthcare company, a team of 10 software engineers have been replaced with three person unit overseeing AI agents. At Amazon, layers of middle management are being stripped out as part of a broader push toward a leaner, AI-ready structure.

AI isn’t just a new tool for the modern workplace; it’s quietly reshaping how companies are organized.

Call it “the Great Flattening.”

As business leaders race to integrate AI across their operations, entry-level roles are disappearing, management layers are thinning, and traditional team roles are starting to blur. Across the Fortune 500, middle management is taking hits, as are entry-level workers, but even at the C-suite, new power dynamics are at play as the old pyramid structure of corporate life starts to flatten out.

Tech bosses have been keen to promote a vision where AI automates the drudgery of work, cutting out admin while allowing soft skills and creativity to flourish. Or, as Microsoft’s Satya Nadella put it earlier this year: “I think with AI and work with my colleagues.”

While the utopian idea of a world without tasks like editing Excel spreadsheets or sorting through files sounds great in theory, what does an AI-first organization actually look like in practice?

An AI flattening

One key theme of organizations that are pivoting to “AI-first” structures is a kind of “flattening” of company structure, which essentially means fewer layers of management oversight, the removal of junior or support roles, and a growing reliance on AI systems to handle tasks that were once handled by human employees.

It can also mean the collapse or conflation of traditional team structures.

For example, pharmaceutical company Moderna recently merged its technology and human resources departments into a single function, appointing a Chief People and Digital Technology Officer to oversee both teams. The move, according to the Wall Street Journal, was driven in part by the company’s partnership with OpenAI and the company leaning into AI to help handle things like HR support and some junior roles.

At consulting giant McKinsey, the company is deploying thousands of AI agents to support consultants with tasks such as building decks, summarizing research, and verifying the logic of arguments. On top of this, around 40% of the company’s revenue now comes from advising on AI and related technologies.

“If you’re an AI-first organization, you can use these AI agents to essentially do a lot of the execution work of organizations,” Nick South, the managing director and senior partner of Boston Consulting Group, told Fortune. “And when we organize our processes and our delivery processes around this AI native workforce, the role of the humans, then, is different.”

This is partly because the nature of individual job roles will change as tasks get automated by AI tools or agents.

“Our job roles get kind of deconstructed, because some tasks might be taken over by AI and others might be new, so the meaning, or the function of the job changes,” Eva Selenko, a professor of work psychology at Loughborough Business School, said. “You might need less of the one role, but that person will take over another function from another thing.”

This doesn’t mean that entire jobs will be replaced, but it could mean employees’ roles become more diverse and take on tasks outside their normal scope of duties, or even their usual team. Job roles get deconstructed as tasks get automated, and their importance in the organization changes, Selenko said. As a result, strict divisions between teams may start to blur.

Mix all this in with a few AI agents carrying out autonomous work, and the traditional org chart starts to look dramatically different.

“Now we’re moving to this more flat network of human teams supervising AI agents,” Rob Levin, a partner at McKinsey & Company, said.

“In early examples, we’re seeing that a client company that’s building an agent factory supporting multiple business workflows, about 50 to 100 AI agents can be managed by just two or three people,” he said.

In one example, Levin said a healthcare company had replaced a traditional 10-person software development team with much smaller, three-person units. These consist of a product owner, a software engineer who can effectively prompt AI coding tools, and a systems architect who ensures integration with the company’s broader tech ecosystem.

However, these kinds of major structural changes are easier to implement in smaller organizations or startups rather than in larger companies that have more complicated structures.

The plight of the middle manager

One of the ways companies, especially in the tech sector, have tried to simplify and flatten their structures for the AI age is to slash employees at the managerial level. Palantir CEO Alex Karp, for instance, announced on Monday’s earnings call that he intended to cull 500 roles from his 4,100-person staff. he called it a “crazy, efficient revolution.”

Middle managers have especially taken a lot of flak, especially from the likes of Big Tech CEOs such as Andy Jassy, who has said that middle managers can hinder speed, ownership, and innovation at Amazon, particularly in the context of AI-driven organizational change.

Jassy is currently pursuing a flatter company structure at Amazon by increasing the ratio of individual contributors to managers, aiming to remove layers and streamline decision-making.

However, experts told Fortune organizations shouldn’t be counting out middle managers just yet.

“One obvious possibility [of flatter organizations] is that it’s going to cause some sort of managerial thinning,” said Tristan L. Botelho, an associate professor of organizational behavior at Yale School of Management. “If AI is reducing the coordination burden, it could shrink the role of middle management, whose job it was previously to make these connections.”

However, while AI might change how middle managers work, Botelho does not expect them to disappear completely.

“I don’t think middle management is going to be erased. I think it’s going to just redefine how managers think about their role within the organization,” he said. “One thing I often talk about a lot with executives…is this kind of idea that AI integrated with the organization, should…level up your skill set as a manager.”

In the AI age, soft skills also become increasingly important, with middle managers serving vital HR functions. Employees still need to be managed, and a workforce still needs to function with empathy or organizations risk losing their best performers.

“There is a human side of things,” Stella Pachidi, a Senior Lecturer in Technology and Work at King’s Business School, said. “It’s not sustainable to just have a boss as an algorithm, it’s not going to work in the longer term.”

In contrast to the Big Tech narrative, some experts say the managerial class will expand within traditional organizational structures as automation replaces low-level work.

BCG’s Nick South told Fortune that execution-level jobs, which are typically at the bottom of the org chart, will be the first on the chopping block due to AI agents, while the managerial or “orchestration” level will grow in complexity and importance.

“At the orchestration layer, that will need to be bigger than it is today…there’ll be a critical human part of that, which is about making sure that all this stuff is doing things,” he said.

“If you think about a classic middle manager profile, it’s sort of a general manager skill set. But what these people are going to need in the future is a combination of sufficient AI proficiency to manage a human-agentic workforce, plus the core skills of logic, understanding of ethics, rhetoric, and communication skills, so that they can communicate with others in much less siloed organizations,” he said. “So those orchestration roles are going to be quite complex.”

The new C-suite

While middle managers and entry-level employees may be feeling the brunt of the AI burden, these changes go right to the top.

AI is already shifting the power dynamics in the C-suite and creating new, powerful, executive-level jobs. According to a 2023 Foundry study on AI Priorities, 11% of mid- to large-sized companies have already appointed someone to serve as “Chief AI Officer” (CAIO), while an additional 21% are currently in the process of recruiting for the role.

According to LinkedIn’s 2023 report on AI at work, companies with a “Head of AI” position around the world had more than tripled in the five years, growing by 13% from 2022.

Alex Connock, a senior fellow at the University of Oxford’s Said Business School, has observed the growth of these roles firsthand.

“Whilst it was perhaps relatively rare when we first launched our AI for business programmes a few years ago, we now have many people … on our executive courses with the title of Chief AI Officer,” he told Fortune. “It’s the new mainstream. The level of interest transcends all levels, from the 20-year-old undergrads … up to seasoned executives.”

While there’s some debate about whether these roles will stand the test of time, for example, experts have said that these roles can lack a clear purpose or authority, the executive level is not immune to the AI boom.

South said that while the rise of some of these roles could also be a symptom of our immaturity with AI rather than a direct need for this work, the C-suite will still have to take on new responsibilities in the AI age.

“People feel quite nervous about missing the boat,” he said. “I think it will be interesting to see how this evolves over time … as things settle down, how different will that be to the classic chief data officer?”

But here is a new “responsibility in the C-suite … to think about our competitive landscape, where are the potential sources of disruption, how we protect our sources of advantage. These are C-suite questions, and it would be a mistake to think, if we hire a chief AI officer, somehow we leave that with them. This is a really top-level thing to be thinking strategically about.”

While AI is set to disrupt all levels of organizations, Selenko noted that management structures are likely to protect those at the top from too much turmoil.

“In an influential management structure, if those at the top’s job might be partly doable by an AI, they will not give up that power. So I think you will see this shift in power balance is probably more among less powerful positions in an organization,” she said.