In order to find AI talent, you need to know where it lives, what skills it has, and how to reach it.
AI hype may grab the headlines, but the real work is figuring out how to build the right skills in the right places. That’s why Lightcast recently published Beyond the Buzz: Developing the AI Skills Employers Actually Need - a research report designed to cut through the noise and identify the skills shaping the future of work.
But while Beyond The Buzz focused on the company perspective, the future of work is as much about people as it is about businesses. That’s where this new three-part series comes in. Building AI Capability: People, Pathways, Places explores the supply side of AI - starting with the talent itself and how to connect with it.
Defining AI Talent: the first step to finding it
The first step in finding AI talent is defining it clearly. Here, AI talent is identified as any worker who holds an AI job title or lists at least one AI skill in their online professional profile, even without having an AI job title.
This methodology aligns with how AI jobs were defined in Beyond the Buzz and in partner research with organisations such as Stanford University’s 2025 AI Index, and the Brookings Institute’s Mapping the AI economy.
This analysis draws on Lightcast’s proprietary dataset of over 800 million online resumes and professional profiles, capturing how workers describe their own skills, experience, and career paths. While coverage varies by country, it offers one of the richest starting points for mapping the global supply of AI talent.
The choice of a skill-based approach opens up a much wider pool than focusing only on job titles, helping uncover talent in unexpected places. AI talent isn’t just one specific group - they span a spectrum of roles, far beyond just engineers. Worldwide, in fact, only 2% of workers with AI skills are AI engineers (around 280,000 people). Another 32% are other tech workers with AI skills (about 3.6 million), and the majority - 66% - are in non-tech roles that still list at least one AI skill (around 7.5 million).
This matters because it opens up opportunities to find AI capability in unexpected places, whether you’re hiring directly into AI roles or looking for people who can apply these skills in other job functions.
The global map of AI talent
AI talent isn’t spread evenly: of the 11.4 million workers listing AI skills in their profiles, the US alone accounts for 35% of them. The US, India, the UK, Canada, and Germany, together make up over two-thirds of the total. But concentration tells a different story: Singapore and Ireland lead the world for AI talent density, followed by Canada, the US, and Luxembourg. Both total numbers and concentration matter and should guide where to look first - large markets offer scale, while high-density hubs can be hotspots for specialised skills and faster hiring.
Where AI engineers live
Broad trends tell you where AI talent is - but it requires more nuance to hire for specialist skills like robotics or deep learning, or to understand AI talent availability across different job functions beyond tech - think marketing or finance.
Zooming in on key AI engineers, the same five countries - the US, India, the UK, Canada, and Germany - account for 60% of the global total. But while the US and India lead across the board, other countries show specific specialties below them. Canada, for example, ranks third in terms of robotics, while the UK ranks third for natural language processing, machine learning, and deep learning.
Availability doesn’t equal easy hiring
Finding talent is one thing, hiring them is another. That’s where it is essential to triangulate multiple data sources. The Lightcast Hiring Difficulty score does this by combining talent availability (profiles), competition (recruitment demand), and compensation to assess hiring difficulties. The score is on a scale from 0 (easier to hire) to 5 (harder to hire).
Let’s continue with the example of key AI engineering roles.
Singapore - the country with the highest concentration of AI talent - is also one where it is hardest to hire AI talent, scoring above four for hiring difficulty for NLP, machine learning and deep learning.
India and the US are also interesting comparisons - both markets have an abundance of AI talent, but because of competition and compensation, it is harder to hire talent in the US than in India.
This underscores the importance of triangulating multiple datasets to refine talent acquisition strategies.
From insights to action - turning data into hires
This blog started broad and then dived into AI engineering talent - but the long tail of AI knowledge across all disciplines is coming fast, without requiring sophisticated coding capabilities.
Knowing where AI talent is - and what skills they bring - can expand the candidate pool beyond job titles. Skills act as signals, helping identify potential hires who might otherwise be missed.
The next step is outreach. With Lightcast’s recent acquisition of Rhetorik, our profiles data has been expanded and enriched to the point that we can provide direct access to a global database of AI talent - making it possible to connect directly with the right people, wherever they are. From hype to hire, the path is now clearer than ever. And it starts with Lightcast.
