Large Market, Unclear Effects on Employment
The global AI market size was valued at $150.2 billion in 2023 and is expected to grow at a CAGR of 36.8% from 2023 to 2030 with a revenue forecast to reach $1,345.2 billion by the end of the period. As AI continues its fast-paced march, questions arise regarding its potential impact on labour markets.
From a theoretical perspective, the net impact of AI on employment is ambiguous: AI will displace some workers but it can also raise labour demand because it supports workers in doing a better, more efficient job, lowering production costs and increasing productivity. AI can also create new jobs since it requires new tasks particularly for workers with skills that are complementary to AI.
Empirical studies in this area find little evidence so far of significant negative employment effects due to AI. This may be due to the fact that AI adoption rates remain relatively low or that firms are reluctant to lay off workers in the short term and/or they may need time to implement new technologies after adoption.
AI Impacts on the Labour Market
Looking ahead, the World Economic Forum forecasts that by 2025, AI could lead to the displacement of about 85 million jobs globally as well as the creation of 97 million new jobs simultaneously. These numbers support other studies that find that most jobs and industries are only partly exposed to automation and are more likely to be complemented rather than substituted by AI.
Varying Impacts by Occupation and by Geography
[1] While the data reported above show outcomes at the global level, it is critical to understand how exposure to AI varies across and within countries. A recent ILO study finds substantial cross-country disparities where 0.4% of total employment in low-income countries is exposed to AI effects, whereas in high-income countries the share rises to 5.5%. This might be primarily due to the different employment compositions between the two groups of countries with high-income countries characterised by larger proportions of high-skill occupations that perform cognitive-based tasks.
A case in point is the United Arab Emirates where, using Whiteshield Labour Navigator we find close to 1 out of 3 AI job displacement risk in occupations such as cleaners and helpers, building structure cleaners, etc.
[1] Source: World Economic Forum, Goldman Sachs, LinkedIn (Future of Work Report), Whiteshield Analysis
Figure 2. UAE – top occupations facing AI job cannibalism
| ISCO code | Occupations | AI job cannibalisation | Share of jobs* |
|---|---|---|---|
| 9112 | Cleaners and Helpers in Offices, Hotels and Other Establishments | 30.6 | 0.40% |
| 1113 | Traditional Chiefs and Heads of Villages | 29.3 | 0.00% |
| 7133 | Building Structure Cleaners | 28.1 | 0.00% |
| 8157 | Laundry Machine Operators | 23.1 | 0.10% |
| 7213 | Sheet Metal Workers | 22.9 | 0.00% |
| 2522 | Systems Administrators | 22.3 | 1.10% |
| 3339 | Business Services Agents Not Elsewhere Classified | 22 | 0.50% |
| 7115 | Carpenters and Joiners | 22 | 0.20% |
| 2656 | Announcers on Radio, Television and Other Media | 22 | 0.00% |
| 7318 | Handicraft Workers in Textile, Leather and Related Materials | 22 | 0.00% |
| 1111 | Legislators | 21.9 | 0.10% |
| 7522 | Cabinet Makers and Related Workers | 21.3 | 0.00% |
| 7215 | Metal Moulders and Coremakers | 21.1 | 0.00% |
| 1341 | Childcare Service Managers | 20.8 | 0.00% |
| 7536 | Shoemakers and Related Workers | 20.6 | 0.00% |
| 5151 | Cleaning and Housekeeping Supervisors in Offices, Hotels and Other Establishments | 20.3 | 0.50% |
| 9311 | Mining and Quarrying Labourers | 20 | 0.00% |
| 1311 | Agricultural and Forestry Production Managers | 20 | 0.00% |
| 4211 | Bank Tellers and Related Clerks | 20 | 0.00% |
| 1420 | Retail and Wholesale Trade Managers | 19.9 | 0.60% |
| 1221 | Sales and Marketing Managers | 19.8 | 6.80% |
| 9613 | Sweepers and Related Labourers | 19.7 | 0.00% |
| 7516 | Tobacco Preparers and Tobacco Products Makers | 19.5 | 0.00% |
| 5152 | Domestic Housekeepers | 19.2 | 1.10% |
| 1219 | Business Services and Administration Managers Not Elsewhere Classified | 19 | 1.30% |
| 3214 | Medical and Dental Prosthetic Technicians | 19 | 0.00% |
The Infrastructure advantage for AI adoption
But different cross-country impacts of AI on labour markets may arise not only from heterogeneous occupational structures, or how well workers are equipped and prepared to work efficiently and effectively side-by-side with machines; importantly, such differences may be driven by the unequal access to AI and data. Whiteshield analysis finds a high correlation between a country’s labour market adaptability to AI and its data and digital infrastructure preparedness (Figure 2). Although many of the emerging economies are already using basic AI technologies (for example in smart farming, credit scoring…), advanced AI technologies are not yet widely adopted despite the tremendous opportunities for economic development that they present.
Figure 3. Affinity of job market AI resilience to infrastructure readiness
Adapted Policies for a Positive Future Impact
Empirical studies also find AI to have a gender-bias impact on women. This might be due to existing gender inequities in professional AI roles globally, over-representation in jobs highly exposed to digitisation as well as gender bias in the technology itself. With the right policy mix, AI offers an opportunity to redress this and other socio-economic challenges including privacy, security, public trust, algorithmic biases, and ethical use of AI.
There is a suite of policies that could be harnessed to amplify the productivity gains that AI promises by focusing on interventions that improve workers’ skills and adaptability, rather than on debating the prospects of machines replacing workers. This calls for a review of education and skills policies to achieve “AI literacy” at different levels of formal education, including in schools to ensure that workers are ready and able to complement AI systems. While specialised AI skills will be required, the shift in skills’ needs is much broader to power occupations such as data analysts and scientists, software and applications developers, and ecommerce and social media specialists, roles that are significantly based on or enhanced by technology. As AI becomes widespread, it will be increasingly important for workers in various occupations to possess a broad range of cognitive and transversal skills to effectively interact with AI systems. Public policy should also promote greater training provisions by employers to ensure life-time learning.