In 2025, 3.2% of job postings in the UAE required AI-related skills according to PwC, while some of the positions created were filled by expatriates. The United Arab Emirates is deploying AI at a pace few countries match, but some of the beneficiaries come from abroad. Singapore conducted its educational reform in parallel with technological deployment, which may influence how the benefits of the transition are distributed. The Emirates can still modify this sequence.
The essentials
- In 2025, 3.2% of job postings in the UAE required AI skills according to PwC, with some positions filled by expatriates.
- PwC indicates that roles most exposed to AI require an average of 77 new skills, a sign of training demand: technological adoption preceded educational reform.
- Singapore pursued training and deployment simultaneously; its system of digital technology training produces cohorts trained locally entering a market where these tools are deployed.
- The adoption-training sequence can influence how productivity gains are distributed between nationals and imported talent.
- A curriculum misalignment can lead an economy to create jobs for which it does not train its workforce.
84% adoption, but the gains go elsewhere
The United Arab Emirates has an explicit and well-funded digital strategy. The country created the world’s first Ministry of Artificial Intelligence in 2017, launched the National AI Strategy 2031, and made technological adoption an explicit state priority. The results are visible: in 2025, 3.2% of job postings required AI skills according to PwC, reflecting growing technological demand.
The 2026 PwC benchmark finds that roles most exposed to AI require an average of 77 new skills, indicating training demand linked to technological adoption. When adopting AI, employers primarily report training their own staff or purchasing external services; recruiting new employees is a less frequent response. Some AI positions are filled by expatriates. Productivity gains are real, but they compensate workers who will return home, save in other currencies, and build their careers in other countries.
This mechanism appears in all economies that adopted a technology without synchronizing their educational apparatus: African countries importing engineers for their digital infrastructure, Gulf states that built their petrochemistry on an almost entirely foreign workforce. AI amplifies this trend.
The Singapore lesson applied to the UAE
Singapore has roughly the same level of technological ambition as the Emirates. Its economy is open, its government interventionist, and its digital strategy dates to 2014 with the Smart Nation Initiative. But the country integrated digital and AI content into education during its technological deployment, an approach the UAE has pursued differently.
In 2024, Singapore’s authorities describe digital and AI education with elective “AI for Fun” modules in secondary school and specialized training in higher education. Singapore’s system of digital technology training, developed in parallel with technological deployment, produces graduates integrating directly into the job market. This figure measures young people trained locally, entering a job market where the tools they learned are actually deployed.
The gap between the two countries does not lie in wealth, ambition, or adoption speed. It resides in the timeline of integration between technological deployment and curriculum reform. Singapore combines training, worker support, and technological deployment. In the UAE, the official strategy treats AI deployment, talent development, and training jointly, without demonstrated comparative hierarchy.
The UAE job market facing its own logic
To understand why the UAE sequence developed this way, and to avoid easy conclusions, context matters. The Emirates have an economy structurally dependent on expatriates since their takeoff in the 1970s, with a significant share of foreign workers in the active population. This dependency predates AI and has never been an obstacle to growth. It has even been a lever: importing skills allows acceleration without waiting for local training.
The problem that AI poses differs from that posed by building skyscrapers or oil extraction. Positions created by AI are analytical, supervisory, design positions—high-value-added, stable positions that build careers. Construction or personal service positions held by expatriates had little effect on the career trajectory of nationals, because these were not positions nationals sought. AI positions, conversely, are positions that the UAE’s Emiratization policy, aimed at increasing employment of Emirati citizens in the private sector, seeks to open to its population, particularly in technology fields.
Emiratization sets local employment quotas in the private sector. The Nafis program, launched in 2021, provides subsidies to companies that hire nationals. These mechanisms work partially: the share of Emiratis in the private sector has grown since 2021. But they hit a bottleneck that quotas alone cannot resolve: absent sufficient local skills, the quota risks becoming an administrative constraint rather than a transformation lever. Technology professions amplify this challenge.
School is the only knot that cannot be outsourced
An employer can train its teams internally. Bootcamps and certificates can fill short-term gaps. The UAE has indeed invested in these mechanisms: Mohamed Bin Zayed University for Artificial Intelligence, inaugurated in 2020 in Abu Dhabi, is the world’s first university entirely dedicated to AI. It trains high-level researchers and engineers. This is a serious commitment, and it will produce results.
But a world-class university alone cannot solve the scale problem between its training numbers and the size of the job market. The gap between the two is structural. What closes it, or fails to close it, is the primary and secondary education system, where entire cohorts that will enter the job market in ten to fifteen years are trained.
Singapore integrates AI literacy into the curriculum, activities, and self-training resources, with strengthened programs in primary and secondary school. This approach takes five to ten years to show up in employment statistics, but it produces effects at the scale of an entire workforce. The article on AI in Africa facing the electricity barrier illustrated a different but similarly structural obstacle: there, infrastructure is missing before training is even a question; in the UAE, technological infrastructure is present while the UAE does have institutional integration of digital and AI content in initial training, notably through a national K-12 framework and prior mechanisms linked to curriculum.
2030: Two trajectories for one economy
PwC’s analysis shows rising demand for AI skills and a broadening of required competencies, without measuring the lag between deployment and training.
A reform of secondary and higher education programs integrating AI tools across multiple fields, with certifications recognized by employers and explicit alignment between competencies produced by education and competencies identified as in short supply, would have the potential to shift the distribution of AI positions between local workers and expatriates. Productivity gains would begin to take root in the national workforce, and Emiratization would become a natural outcome rather than an administrative constraint.
Maintaining technological deployment without parallel educational reform risks consolidating dependence on expatriates in the most skilled positions. Companies would accept this because it would be functionally advantageous in the short term. The UAE economy would finance in this scenario a social ascent benefiting partly workers trained abroad.
This second scenario is not an economic catastrophe in the short term. The UAE would remain a dynamic, attractive, well-managed economy. But it represents a considerable opportunity cost for local generations entering the job market in 2025-2030 in an economy saturated with tools they have not learned to master. The social mobility promised by the 2031 AI Strategy would remain theoretical for a fraction of the national workforce.
Two signals will allow distinguishing trajectories before results become visible. First: the annual evolution of the share of AI positions filled by nationals in organizations that have adopted AI. Sustained growth would signal that curriculum reform is producing its first effects; stagnation would indicate the gap is widening. Second: the average lag between a technology’s introduction in businesses and its appearance in initial training programs. Both indicators are currently measurable in the UAE and would allow comparing trajectories annually.
What a curriculum monitoring institution would change
Part of the UAE’s difficulty lies in an institutional gap. The UAE has recently established a structural mechanism explicitly linking educational data, demanded skills, and employer needs; the historical frequency of such alignments is not established. The PwC benchmark partially fills this void, but it is an annual consulting firm report, not a public steering mechanism.
Several advanced economies have created this type of institution. Singapore has SkillsFuture Singapore, which develops with employers and partners competency frameworks and training information to meet sectoral needs. France has its network of sectoral observatories, imperfect but active. Germany has articulated dual training with business needs for decades. These models are transferable: they require no particular culture, but political will and a budget.
Such a UAE institution would have a precise role: calculate in real time the gap between skills employers declare as in short supply and programs in place in secondary schools and universities, then trigger curriculum revisions within a set timeframe. This approach exists in other contexts: Singapore’s EdTech Masterplan 2030 establishes connections between technological evolution and adaptation of school programs.
The Emirates have shown on other topics that they know how to create new institutions quickly when political will is present: the Ministry of AI is the most direct example. Applying this same will to training, a domain less visible than technological deployment, could influence how benefits from deployment are distributed. Analysis of who captures gains from a technological transition in global value chains shows that benefit follows the skills, institutions, and political decisions that produce them.
The sequence as public policy
PwC documents rapid transformation of AI-related skills. A gap between technological deployment and local training can increase reliance on foreign talent and direct part of productivity gains toward a mobile workforce.
Economies of very different sizes have adopted a common approach: adapt schools not to today’s needs but to tools students will find when entering the job market in ten years. Singapore has gradually developed its digital education. Estonia integrated coding and digital tools into primary school starting in 2012, anticipating future job market demands and producing a tech-dense workforce.
The UAE has the financial resources, displayed political will, and a territory small enough to conduct educational reform quickly. Treating training as a structuring element of deployment, rather than as its natural extension, remains the decision to be made. Explicitly linking public AI deployment authorizations to simultaneous implementation of certifying training programs for affected local workers would constitute a clear signal: a definition of what it means to adopt AI for an entire economy, not just its organizations.
Sources
- PwC AI Jobs Barometer UAE 2026
- Singapore Ministry of Education, EdTech Masterplan 2030 and Smart Nation 2.0 (official publications)
- Consultancy Middle East, analysis of the AI job market in the UAE, 2026
- IMF / World Bank, reports on the education-employment gap in the MENA region
- SkillsFuture Singapore, annual reports on skills in short supply (skillsfuture.gov.sg)
