AI and the economy: the great divider

Artificial intelligence (AI) is not just another productivity tool quietly improving efficiency. It is a dividing force - one that will reshape economic outcomes by accelerating disruption and redistributing power across firms, workers and nations.

The question is no longer whether AI will transform the economy (it will and already is) but how can we ensure a smooth transition and prevent mass economic dislocation tearing apart the foundations of our societies.

Increased automation has long been a defining feature of the evolution of many professional sectors. However, unlike previous waves of automation, AI does not simply just replace routine labour. It reshapes decision-making, compresses time, and alters the economics of scale itself.

In doing so, it creates clear winners and losers faster than markets, societies, and - most definitely - governments are able to respond. 

Large, well-capitalised firms 

The clearest winners from AI adoption will be large firms with scale, data and capital.  

AI thrives on three things: data volume, computing power and integration across complex systems. Big firms already possess these advantages. 

These companies can absorb the high upfront costs of AI infrastructure; train proprietary models using vast internal datasets; and spread AI investment across multiple business lines. 

As a result, AI allows large firms to operate faster, cheaper and more precisely than smaller rivals. 

Market concentration is likely to increase in sectors where AI delivers strong economies of scale, raising long-term questions about competition. 

High-skill, high-judgement workers 

AI disproportionately benefits workers who combine specialist knowledge, judgement and creativity - senior professionals, technical leaders, strategists and hybrid roles that sit between technology and decision-making, as well as skilled tradespeople. 

These individuals will harness the soft skills that set them apart, becoming more productive, more valuable, and harder to replace. AI acts as a force multiplier for their output. 

Wage dispersion will increase as top performers gain leverage and technical, manual skills become more valued, while mid-level, task-oriented roles face pressure. 

Owners of infrastructure, data and energy 

AI is capital-hungry. Data centres, cloud platforms, chip manufacturers, grid operators and energy providers sit at the heart of the AI economy. 

As demand for computing and power surges, these assets gain strategic importance. Increasingly, those who already control infrastructure will retain and grow their economic influence. 

The symbiotic relationship of labour and capital will be upended, with returns shifting towards capital and infrastructure ownership, reinforcing long-term wealth concentration. 

Governments that act early and strategically 

Yet countries that invest in digital infrastructure, skills, energy capacity and AI governance will attract capital and talent. Those that align regulation with innovation - cautious but supportive - will shape global standards. 

Not all countries will get this right. AI may widen gaps not just between firms, but between nations. 

Firms that treat AI as a bolt-on 

Businesses that view AI as a standalone IT upgrade - rather than an organisational transformation - risk falling behind. Without process redesign, cultural change and leadership commitment, AI adds complexity instead of productivity. 

These firms may see rising costs without proportional gains. Poor integration may lead to slower decision-making, meaning they lose ground against AI-native rivals. 

For companies that don’t effectively integrate AI into their businesses, the gap between themselves and AI leaders will widen rapidly. 

Middle-skill, task-based knowledge workers 

The most exposed group in the labour market is not low-skill manual workers, but middle-skill cognitive roles: analysts, administrators, junior professionals and support functions built around information processing. 

AI can perform many of these tasks faster and cheaper. While new roles will emerge, transitions will be uneven, and economic dislocation will occur. For graduates looking to enter these jobs, opportunities may become scarce. 

Without large-scale reskilling, this group risks wage stagnation, displacement or downward mobility. Without an overhaul of the education system, many in the next generation may struggle to even enter the workforce. 

Small firms without access to capital or data 

Small and medium-sized enterprises face a structural challenge. AI lowers marginal costs but raises fixed costs. Firms without access to capital, data or specialised talent may struggle to adopt AI effectively. 

Even where demand remains strong, competitive pressure intensifies as larger rivals operate at lower cost and higher speed. Entrepreneurship may become harder in some sectors, not easier. 

Governments that lag behind 

Governments have rarely kept pace with technological change. In the 21st century, social media and data monetisation have consistently raced ahead of regulation, leaving policymakers in a permanent state of catchup. 

States that fail to regulate effectively, modernise their public services or invest in skills, risk falling into a vicious cycle: weaker growth, lower tax revenues and rising social pressures. 

Public institutions that cannot match private sector AI capability may struggle to deliver services efficiently or regulate credibly. 

Society at large: uneven gains and rising tensions 

At a societal level, AI creates a paradox.  

Technology has often been considered a great leveller, enabling anyone with access the ability to perform tasks to a high level.  While AI can raise overall productivity and wealth, it will simultaneously deepen inequality and social tensions. 

If AI-driven gains accrue mainly to capital owners, top performers and large firms, public resistance will grow - even if aggregate economic output improves. Trust, fairness and transparency become central to economic sustainability. 

So, what happens to consumers? If capital, employment and ultimately wealth increasingly concentrates at the top, eroding consumers’ spending power, how does our demand-driven economy survive? 

Does greater productivity and efficiency even matter if there are too few people able to buy what is produced? 

AI does not determine outcomes on its own. Policy, strategy and governance matter. Governments can shape labour transitions, competition frameworks and education systems. Businesses can choose between long-term value creation and short-term cost cutting. But who funds the skills transition - businesses or the state? 

AI creates a two-speed economy: 

• fast, AI-enabled firms and a small number of workers pulling ahead 

• slower, under-capitalised segments falling behind. 

Once established, this divide becomes self-reinforcing.

AI is not simply changing how work is done; it is filtering the economy - rewarding preparedness, scale and adaptability while penalising inertia. 

The biggest winners will be businesses with access to technology and capital, and individuals with the human skills to apply them effectively.  

The greatest losers will be businesses that lack capital or fail to adapt their institutions, skills and incentives quickly enough, and knowledge-based workers who do not adapt alongside them. 

This is why the defining challenge of the AI era is not technological, but economic, political and social. 

AI is changing our world. 

Can we harness it to broaden prosperity, or will we allow it to concentrate power and opportunity even further? 

About the author   

Emeritus professor of global economy at Cranfield School of Management, Joe Nellis CBE is one of the UK’s most experienced and well-known economists, with four decades of experience commenting on UK, European and global macro-economic trends.   

Joe is a frequent commentator to national print and broadcast media on issues such as public investment, GDP and growth, tax and the wider economy, as well as data points such as inflation, unemployment and interest rates.  

He has published 19 research and subject-based books and over 200 academic and practitioner journal articles. His research encompasses analysis of business developments in a changing world in terms of the macroeconomy, the role of government, the impact of technology, societal and demographic trends.