CAMBRIDGE – Many fear that AI will cause “job destruction”. While this year’s Davos conference sounded the alarm about the impact of technology on jobs, recent announcements about layoffs in white-collar industries have been widely seen as a straw in the wind.
However, the far-reaching impact of AI on business has rarely received enough attention. The most reliable research shows that the majority of companies have not yet adopted AI, and that future adoption of AI is likely to involve large-scale corporate restructuring. That’s because AI is an information technology that influences decision-making processes.
The wave of digital technology since the 1990s has transformed business in many ways. Advances in computers and communications powered the Internet, and the advent of smartphones and wireless network technology made the Internet mobile. These have enabled the transition from vertically integrated production to globally distributed supply chains and from corporate hierarchies to “backward” organizations. Indeed, policy and regulatory changes have facilitated the globalization of production and significant growth in cross-border trade in components. However, these changes would not have been possible without technological innovation.
Another impact of digitalization is the rise of platform business models that use algorithmic tools to mediate between suppliers and customers and build extensive logistics networks on top of digital infrastructure. Data and algorithm-driven platforms operate in many sectors, often dominating markets, and have transformed both employment and consumption patterns.
The question now is how AI will rewire enterprises. At the World Intellectual Property Organization last summer, Vivek Mohindra, senior vice president and special advisor to the vice chairman and COO of Dell Technologies, argued that “organizational capabilities” are a source of a company’s sustainable competitive advantage (Dell’s key intangible asset is its supply chain). But AI is changing key capabilities, making them difficult to measure, he added.
Some industries appear to be particularly vulnerable to disruption by AI. Several commentators have already pointed to the technology’s potential to automate entry-level jobs in fields such as law, accounting and finance. Similarly, tech companies are increasingly using proprietary AI models to reduce software development time and costs, suggesting that fewer computer programmers will be needed in the future.
But once a company’s bottom ranks are decimated, how can companies ensure that future employees have the expertise they need? For example, there is emerging evidence that using AI to write code undermines skill acquisition in human workers.
Generative AI will also reshape corporate structures. One possible outcome is that technology will continue the process of flattening organizations, allowing them to outsource more and more jobs. OpenAI’s Sam Altman even predicts the possibility of one “unicorn” (a billion-dollar startup). AI agents can reduce the friction inherent in negotiations between different actors and monitor complex supply chains.
However, some economists predict a recentralization of organizations because of generative AI’s ability to capture the “tacit” knowledge embedded in human cognition and practices, the knowledge that all businesses rely on.
Let’s consider a small example. A maintenance engineer working on the London Underground noticed that the wheels on Victoria Line cars needed extra grease because of the unusual bends in the track. When these employees left, that know-how disappeared, and Victoria Line trains often broke down due to worn wheels.
Tacit knowledge, such as that of maintenance engineers, is rarely documented or formally taught. But if it reflects the repeated actions of human workers, new AI applications could capture and codify this know-how.
Business and political leaders need to track AI-driven organizational changes as they occur to be better prepared to respond to the structural changes that seem inevitable. A key part of this effort is to enable individuals to weather large-scale labor market disruptions more easily and better than in previous waves of automation. Governments should take special care to avoid a repeat of the inadequate policy response to manufacturing automation in the 1990s that left post-industrial scars in many developed countries.
Being prepared will also help companies deploy AI in a way that strengthens organizational capabilities and, in turn, improves productivity across the economy. All business leaders need to consider how generative AI can be integrated into their organizations, particularly how it can transform production processes, what tacit knowledge can be harnessed more effectively, and who should be responsible for decision-making. That means you need a plan to keep your company competitive in this new wave of technology.
The structural changes that the economy will experience over the next decade are almost certain to be as dramatic, if not more dramatic, than recent changes. As many now expect, AI will transform work. But we must remember that it will also reshape the business landscape.
Diane Coyle, professor of public policy at the University of Cambridge, is the author of Cogs and Monsters: What Economy Is, and What It Should Be (Princeton University Press, 2021) and The Measure of Progress: Counting What Really Matters (Princeton University Press, 2025). This article was distributed by Project Syndicate.
