Global AI: Who’s Leading the Race for Jobs and Skills

Demand for artificial intelligence skills has boomed 33%, on average, across 14 countries in North America, Europe, and the Pacific, according to a new joint OECD-Lightcast report released today. The report, based on Lightcast job posting data, offers a pioneering global view of demand for AI jobs and skills from 2019 to 2022. 

While AI is a global phenomenon, it’s not going to play out in the same way everywhere. Different countries have their own industries, investment strategies, and workforce skills. While the US has the most AI job postings overall, other countries are already starting to build their own areas of expertise, such as autonomous vehicles in France and robotics in the Netherlands. 

To keep things in perspective, the overall proportion of AI job postings remained small compared to the total market. That ranged from 0.84% in the US to 0.11% in New Zealand. But those roles, like AI in general, are going to have a disproportionate impact on the job market.

Job postings provide a unique perspective into emerging technologies, because this is where employers show their hand. Companies have to specify the skills they want and need. So if the world is in a race to develop artificial intelligence, analyzing jobs and skills tells us who’s ahead, who is gaining, and what technologies matter.

AI Demand Across Countries

There are two major signs of a vibrant AI skills market: how much employers are clamoring for these emerging skills, and how fast that demand is increasing. The countries examined include some English-speaking nations (Australia, Canada, New Zealand, the United Kingdom, and the United States) as well as selected European countries where Lightcast has deep resources of data (Austria, Belgium, France, Germany, Italy, the Netherlands, Spain, Sweden, and Switzerland).

If you look at where the jobs are, the US is leading with the largest share of artificial intelligence roles in job postings, both in 2019 and in 2022. The US also had healthy growth, as AI postings grew 22% in the period.

But while other countries may have had smaller numbers of AI postings, their postings have grown at a faster rate. Spain, for example, started out with a small number of AI postings, ranking 11 out of the 14 countries studied in 2019. But Spain saw dramatic growth (155%) and by 2022 was in fifth place. 

This suggests a rough shape for the emerging geography of AI: there will certainly be countries that are powerhouses in the field, but there could also be pockets of specialized innovation, where countries build strength in a specific field.

AI Occupations: Still Looking for Developers

There’s been huge interest in how AI will change a wide range of occupations as more and more organizations look to harness the technology. Eventually, that’s going to be reflected in AI skills being included in all kinds of job postings. 

But so far, the main demand for AI is still in the occupations and industries most likely to be building AI applications.

The largest shares of artificial intelligence postings were in Professional Activities (professional, scientific and technical services / activities industry, which includes consultancies), ICT (another term for information technology), and Manufacturing. Equally significant is the fact that the biggest growth in posting was in those industries. Few if any AI postings, and not much growth, happened in areas like Agriculture, Accommodation, or Transportation, the report said.

It’s significant that what is usually known as “the tech industry” is included in Professional Activities and ICT. That suggests that the hiring and demand for skills is still among tech companies developing AI rather than other kinds of employers adopting it.

“A key shift to watch for is whether AI demand spreads outside the ‘big three’ industries,” said Layla O’Kane, Senior Economist at Lightcast and an author of the report. “That will show us when AI starts reshaping jobs on a wide scale, with human workers alongside AI applications.”

But the focus can shift depending on the country, and the strengths of its economy. Fully half of all the AI postings in New Zealand are in Professional Activities, compared to 36% in France and 30% in the US. 

In Spain, Belgium, and Italy more than four in 10 AI postings are in ICT. And manufacturing leads in Germany, Sweden, and Switzerland. This is partly due to the existing local specialization of these economies and partially how far they are in their path of adopting AI. In every country, demand is a combination of new firms focusing on AI (typically in the ICT or Professional Services industries) and existing companies looking to integrate AI into their workflows. Demand for AI in Spain, which saw the most dramatic growth, is mostly in the ICT industry and, with a strong focus on Machine Learning.

AI Skills

Since the launch of ChatGPT and its competitors, generative AI and its potential for upending jobs has been in the news. But other, more established technologies dominate job postings.

Machine Learning is the skill in highest demand across all the countries examined. This is the technology that drives algorithms already built into many applications. Lightcast, for example, has been using machine learning for 20 years as a core element in collecting and analyzing more than 1 billion job postings and career profiles. So despite the tendency to think of AI as brand new, some of the most important skills can be more than two decades old. 

The report used a skill cluster analysis, grouping like skills together. Across the 14 countries examined, roughly one-third (34% ) of AI postings required skills from the Machine Learning cluster, 21% asked for skills in the Artificial Intelligence cluster, and 14% wanted skills in the Neural Networks cluster.

The basic pattern held across all the countries, but the proportions changed–again signaling the relative strengths and priorities in different economies. For example, the demand for machine learning ranged from 39% of AI posts in the US to 29% in France.

Yet other powerful skill sets were more in demand. Autonomous driving, which was only 0.05% of postings across all countries, accounted for 19% in France and 17% in Sweden. Robotics reached 13% of postings in the Netherlands. This further illustrates the point that as AI grows, not everyone will become a leader in all things associated with artificial intelligence, instead we’ll see pockets of expertise arise in distinct places.

Artificial intelligence job postings show us the patterns that are already developing in this critical field. AI is an addition, not a substitution, for more traditional skills. Workers, companies, and even nations will add AI to their existing strengths, attempting to exploit the advantages they already have. While generative AI only burst on the scene in 2023, it is likely that type of AI will follow a similar pattern.

The OECD report provides key insight for all the potential players in the job market:

• Businesses can identify the skills they need for AI development—and also see what their competitors are already doing.

• Governments can scope out global competition while making better local investments in economic development and training.

• Educators can determine which skills are likely to be in demand and create programs to meet those needs.

• Workers and students can better understand the trends in this fast-growing field. For those looking to switch into AI–or to get in at the start of this revolution–getting new skills mapped to what’s going on in their nation will be key to success. 

To learn more about global trends in AI, geopolitics, and labor shortages, read our latest research, Fault Lines.