A woman speaks with her colleagues, while they look on, smiling.

Understanding the Talent Infrastructure Powering AI Transformation

LinkedIn logo

Economic Graph Team

Written by Mar Carpanelli, Akash Kaura, and Rosie Hood

Companies are racing to embrace Generative AI: spanning industries and job functions, businesses are looking for outsourced solutions or building in-house. What began for most as an interest in prompt engineering, has evolved into investment in AI engineers, researchers, and consultants. However, there is more to the AI value chain than model development to help drive AI adoption. Unlocking the full potential of AI requires a diverse set of talent spanning technical, operational, and governance roles. To reflect this, we present a framework that represents the complete AI value chain and lifecycle.

As technologies advance, applications broaden, and, as a consequence, new workforce needs emerge, and we believe an interoperable workforce framework powered by near-real time labor market data is critical to understanding this fast evolving space. Based on analyzing trends from LinkedIn’s 1.2+ billion members, our goal is to provide a shared language for understanding the AI workforce so that educators, policymakers, and businesses can align on long term investments to develop inclusive AI ecosystems.

The AI Value Chain Workforce Framework

The AI Value Chain

A value chain is a way of describing all the different steps, people, and activities needed to turn an idea into something that reaches people. It is like a “chain of links” where each link adds value, like designing, building, shipping, selling, and supporting a product. 

For AI, the value chain includes several layers of talent and a constantly evolving workforce landscape:

  • Inputs and Infrastructure (electricity, semiconductor chips, data centers, broadband): These are necessary inputs required for large scale computation required to develop AI models including Large Language Models (LLMs).

  • Development (AI engineers, machine learning engineers, data scientists): These professionals develop, train, and fine-tune AI models, as well as tools on top of those models to address a variety of business use cases.

  • Trust, Safety, and Governance (Professionals in cybersecurity, ethics, policy): These workers ensure AI development and use are safe, fair, and accountable.

  • Adoption (AI-literate professionals): These professionals use AI models, tools or applications in their day-to-day workflows, driving adoption of AI tools by businesses at scale.

Towards a Workforce Framework

Going from inputs on the left to adoption on the right, we can see the various parts of the value chain provide a useful framework for thinking about this evolving landscape and the workforce that is powering it. Within each of the 4 value chain layers, we see 8 specific categories around which work can be organized.


A chart that explains each of the four AI value chain layers.

In order to better understand the role of each of these talent pools in the AI value chain, here's a closer look at each, including what they do and why they matter, as well as examples of job titles that fall into their respective talent pool.

It must be noted that most of these talent pools in the AI value chain represent a highly specialized and scarce group of in-demand workers. In fact, per our just released AI Labor Market Update, hiring for Head of AI Engineering talent has grown by more than 25% year-over-year in 2025, with AI job postings accounting for 7% of all technical job postings on LinkedIn, even though AI talent only represents less than 1% of LinkedIn members in the United States.

However, the AI literacy professionals category is different and is not limited to a niche workforce. Just as digital literacy became essential in the internet era, AI literacy is likely to become a universal requirement for participation in the modern economy.

Operationalizing the Framework and Next Steps

Data from the Economic Graph provides near-real time updating lists of skills and titles required to define and measure the progress of various industries or geographies in one or more of these categories. Our work on measuring the AI Engineering and AI Literacy talent pools is featured in the Stanford AI Index on an annual basis, and is also a key component of the OECD AI Observatory. These are industry-leading, comprehensive data-driven resources focused on AI, and reveal insights such as the unprecedented growth of the AI Engineering workforce in response to the rapidly accelerating adoption of AI among employers — as of 2024, 7 out of every 1,000 LinkedIn members globally are considered AI engineering talent, a 130% increase since 2016.

Over the coming months, we will share such insights into each component of the AI value chain including showcasing trends unique to each specific area. We look forward to sharing our upcoming reports on Energy and Sustainability in AI and a deep dive on Data Center Operations Talent.