The risks of a two-speed AI economy in MENA

Kamran Yeganegi
September 08, 2026

Much of the debate about artificial intelligence in the Middle East and North Africa focuses on how many jobs it will replace. As this column explains, the more important challenge is that the new technology is likely to create a two-speed labour market, where countries with the capacity to invest in digital infrastructure and skills move rapidly ahead while others struggle to keep pace. Unless governments combine credible national reforms with targeted regional cooperation, AI could widen, not reduce, the economic divides across MENA.

In a nutshell

MENA risks becoming a two-speed AI economy: a regional labour landscape in which differences in preparedness, adoption capacity and access to technology produce divergent trajectories for productivity, wages and employment.

The greatest danger is not only technological unemployment; it is technological marginalisation: workers and firms may remain concentrated in low-productivity activities while investment, digital industries and career mobility increasingly cluster elsewhere.

Across MENA, there are wide disparities in fiscal capacity, digital infrastructure, educational quality, innovation systems and labour market institutions; AI may reinforce those inequalities unless policy addresses both national constraints and cross-border gaps.

The policy debate on artificial intelligence (AI) and employment usually begins with a familiar question: how many jobs will AI replace? For the Middle East and North Africa (MENA), the more consequential issue is not simply the number of jobs that may disappear, but the unequal capacity of economies to adopt AI productively and create new activities around it.

I describe this risk as a two-speed AI economy: a regional labour landscape in which differences in preparedness, adoption capacity and access to technology produce divergent trajectories for productivity, wages and employment.

This is a stylised description of widening divergence – not a strict division between two homogeneous groups of Gulf and non-Gulf economies. There are substantial differences within both groups and, equally importantly, within countries: between large firms and small and medium-sized enterprises (SMEs), capital cities and peripheral regions, highly skilled and less-skilled workers, and formal and informal labour markets.

The distinction matters because MENA already exhibits wide disparities in fiscal capacity, digital infrastructure, educational quality, innovation systems and labour market institutions. AI is therefore arriving in an unequal region and may reinforce those inequalities unless policy addresses both national constraints and cross-border gaps.

AI preparedness already differs sharply

The AI Preparedness Index produced by the International Monetary Fund (IMF) captures four foundations of effective adoption: digital infrastructure; human capital and labour market policies; innovation and economic integration; and regulation and ethics. A recent IMF study reports an index score of 0.58 for Saudi Arabia, compared with 0.48 for Jordan, 0.47 for Tunisia, 0.43 for Morocco and 0.39 for Egypt (Cerutti et al, 2025).

These figures should not be read as a permanent ranking. They do, however, illustrate that countries enter the AI transition with materially different capacities to convert technological exposure into productivity gains.

The same IMF study estimates that the growth impact of AI in advanced economies could be more than twice that in low-income countries. Its central conclusion is directly relevant to MENA: outcomes depend not only on occupational exposure to AI, but also on preparedness and access to essential technologies and data. Better preparedness can narrow the gap, but may not eliminate it.

The Gulf economies enter the AI era with important advantages, including substantial public investment capacity, expanding cloud and data centre infrastructure, national AI strategies and the ability to attract global technology firms and specialised talent. Yet these advantages are not evenly distributed. SMEs, less connected regions and workers outside high-productivity sectors may still face major barriers to adoption.

Many oil-importing economies face tighter fiscal constraints, persistent skills mismatches, weaker firm-level digital capabilities and larger informal labour markets. Nevertheless, they are not a single slow-moving bloc.

For example, Jordan and Tunisia possess significant pools of digitally capable graduates, while Egypt and Morocco have developed export-oriented service and outsourcing activities. Their challenge is less an absence of talent than the difficulty of combining talent with finance, infrastructure, firm capabilities and scalable demand.

From exposure to employment: four transmission channels

Exposure to AI is not the same as adoption, and adoption is not the same as job displacement. Four transmission channels need to be distinguished:

First, occupational exposure measures the share of tasks that AI can potentially affect. A highly exposed occupation may experience substantial change even if the occupation itself does not disappear.
Second, firm capacity determines whether potential exposure becomes actual adoption. Firms need data, computing access, finance, managerial capability, cybersecurity and workers able to reorganise production around AI. A large bank and a small retailer operating in the same country may therefore face entirely different AI transitions.
Third, the effect on workers depends on whether AI substitutes for tasks or complements human capabilities. In some administrative and routine cognitive activities, firms may reduce labour demand. In health, education, engineering and professional services, AI may instead raise worker productivity and expand demand – provided that institutions and business models can translate efficiency gains into wider activity.
Fourth, national production structures determine whether productivity gains generate new investment and employment. Economies able to build new digital services, Arabic language applications and AI-enabled industries will capture more of the value created. Economies that remain consumers of imported systems may gain efficiency while creating comparatively few domestic high-value jobs.

The greatest danger is therefore not only technological unemployment. It is technological marginalisation: workers and firms may remain concentrated in low-productivity activities while investment, digital industries and career mobility increasingly cluster elsewhere.

Comparative advantage should shape the regional response

The solution is not for every MENA country to imitate the Gulf model. Economic history demonstrates that regions prosper when countries specialise according to their comparative advantages rather than pursuing identical development strategies.

Capital-rich economies possess advantages in financing computing infrastructure, frontier research and large-scale technology deployment. Labour-abundant economies can develop competitive capabilities in software engineering, Arabic language AI applications, business process outsourcing, digital health services, educational technologies, creative industries and AI-assisted professional services.

Rather than competing to replicate the same technological ecosystem, MENA economies should build complementary positions within regional value chains. Such an approach could widen the geography of AI-enabled employment while allowing national strategies to reflect different economic structures and institutional capacities.

Inclusion is an economic requirement

A two-speed transition can also emerge within each country. In 2023, 31.5% of young people in MENA were not in employment, education or training; among young women, the rate reached 44.2%. The region’s youth unemployment rate was 24.4% – almost twice the global average. The International Labour Organization (ILO) also reports that the vast majority of young people in paid employment in the region work informally.

These conditions shape who can benefit from AI:

Women may gain from remote and flexible work, but unequal access to connectivity, care responsibilities and concentration in some exposed clerical occupations can limit those gains.
Young graduates may be digitally familiar yet lack the applied and interdisciplinary skills that employers require.
Informal workers are often excluded from employer-funded training and contributory social protection.
SMEs may be unable to afford data systems, advisory services or secure computing even when low-cost AI tools are technically available.

Inclusion should therefore be treated as part of productivity policy, not as an afterthought. Public programmes should combine affordable connectivity, modular digital training, recognition of micro-credentials, SME adoption support and portable access to reskilling and social protection. Programmes should report outcomes by gender, age, firm size, region and employment status so that aggregate progress does not conceal widening internal gaps.

From talent competition to talent circulation

The rapid expansion of AI industries in wealthier economies is likely to increase demand for engineers, data scientists, cybersecurity specialists and digital professionals. Without coordinated policies, this demand could accelerate the permanent migration of highly skilled workers from lower-income MENA countries, deepening existing disparities in human capital.

The region should move from fragmented competition for talent towards talent circulation. Temporary mobility, joint research programmes, mutually recognised professional credentials, cross-border remote work and return incentives can allow knowledge and income to circulate. Mobility should not be restricted: it should be structured so that origin economies, destination economies and workers can all benefit.

A realistic MENA AI labour compact

Current labour market policies remain largely national while technology firms, digital services and skilled labour increasingly operate across borders. I therefore propose a voluntary MENA AI labour compact. This should complement, not replace, national reforms in education, competition, social protection, digital regulation and private sector development.

The compact could begin with a small coalition of willing governments, supported by the ILO and UN Economic and Social Commission for Western Asia, with the ERF serving as a knowledge partner.

Participation should remain voluntary, and implementation should rely on national institutions. This arrangement would avoid creating a costly new regional bureaucracy while providing technical legitimacy, comparable evidence and a platform for coordination.

During its first two years, the compact should deliver four practical outputs:

First, a regional observatory should publish comparable indicators on occupational exposure, firm adoption, skills demand and distributional outcomes.
Second, participating countries should agree on a common competency framework for selected AI-related occupations and recognise a limited set of quality-assured micro-credentials.
Third, two or three pilot partnerships should link finance from capital-rich economies with training institutions and firms in labour-abundant economies, with transparent targets for employment, women’s participation and SME inclusion.
Fourth, participating governments should test portable reskilling support for freelancers, platform workers and informal workers who are commonly excluded from conventional training systems.

Success after two years should be judged by concrete measures: comparable data coverage, credentials recognised, workers trained and placed, SMEs adopting relevant tools and the share of women and disadvantaged workers reached. Only after evaluating these pilots should the compact expand its mandate.

The future divide is between uneven capacities to adopt

The future of work in MENA will not primarily be determined by whether machines replace people. It will be determined by the interaction between occupational exposure, firms’ capacity to adopt AI, the balance between substitution and complementarity, and countries’ capacity to create new activities around intelligent technologies.

Some economies may become global centres for AI investment, innovation and high-productivity employment. Others may remain concentrated in lower-productivity labour markets with persistent skills mismatches and slower technology diffusion. But neither trajectory is predetermined, and the dividing line will not fall neatly between the Gulf and the rest of MENA.

Policy choices made today can narrow the emerging divide. National reforms must strengthen infrastructure, skills, institutions and the capacity of firms to innovate. Regional cooperation can add value where problems are genuinely cross-border: comparable labour market intelligence, portable skills, talent circulation and shared investment.

The real question is not whether AI will reshape MENA labour markets. It almost certainly will. The question is whether the region can turn unequal starting conditions into complementary capabilities – or allow them to harden into permanently unequal futures.

More than 2 million jobs are needed each year to absorb new entrants into Egypt’s labour market and raise the country’s employment rate. The job challenge is even more acute for women, whose labour force participation remains low despite recent gains in education. This column reports on the second Development Dialogue, an ERF–World Bank Group joint initiative, which brought together students, scholars, policy-makers and private sector leaders at the American University in Cairo to consider how the country’s gender gap in work can be closed.

Digitalisation, global value chains and regional integration in MENA & SSA

Participation in global value chains is vital for countries pursuing structural transformation and inclusive economic development. This column summarises new evidence on how much production processes have been globalised in Africa and the Middle East relative to other regions; whether this process has taken place with partners within or outside the region; and whether it has taken place more in manufacturing or services.

How trade policy can reduce MENA’s cereal import vulnerability

Heavy dependence on imported cereals, combined with climate change, water scarcity and geopolitical uncertainty, continues to threaten food resilience across MENA. This column explains how an inclusive trade policy can play a key role in making the region’s food security less vulnerable to shocks.

The political economy of algorithms in the Middle East

The Middle East is entering a new political-economic era – one in which algorithms, data and artificial intelligence may become as strategically important as oil once was. Across the region, governments are investing heavily in digital infrastructure, smart governance and AI-driven economic transformation. This column outlines how AI and algorithmic governance are reshaping power, inequality and state capacity in the region.

The risks of a two-speed AI economy in MENA

From mega-investment to mega-productivity: closing MENA’s conversion gap

Across the Middle East and North Africa, governments are investing heavily in infrastructure, technology, diversification and new industries. Yet capital accumulation alone does not guarantee productivity growth. As this column argues, the region’s deeper challenge is its ‘investment-to-productivity conversion gap’: ensuring that large-scale investment translates consistently into stronger firms, technological diffusion, skills and better jobs. A different metric of success is required: not how much capital is deployed, but how much productive capability that capital leaves behind.

The Forum ERF

ERF