The public debate on AI is one of the most polarized of recent years. On one side are those who warn against its uncontrolled development, some of the world’s leading AI developers among them. On the other are those who see artificial intelligence as a bottomless pool of growth that will make life easier and more convenient.

In between are millions of people who have started using the new technology at work every day. The consulting firm PwC (PricewaterhouseCoopers) surveyed 49,364 of them across 48 countries and territories, between May and June 2026, to understand how AI is reshaping everyday tasks, how the tools are actually used, and what companies stand to gain from them.

The picture that emerges from the survey points to a concrete risk: a widening gap between first-tier and second-tier workers.

PwC’s four categories

PwC’s analysts crossed two variables: how rare a worker’s skills are, and how much benefit that worker says they get from using AI.

Four categories came out of it: the Front-runners, the AI insurgents, the Indispensables and the Engine room.

The first group, the pioneers, accounts for 14% of the sample. They have rare skills and an already settled relationship with AI: 51% use generative AI every day, and almost 80% have access to the training resources they need. They are the ones positioned to get the most out of the technology, and it is no accident that roughly 29% are ready to change employer within a year.

Another 18% of the sample falls into the AI insurgents: workers whose skills are less rare, but who manage to use AI to get more done and close the distance with the pioneers.

A further 11% are the Indispensables. They still make little use of AI, but they hold rare skills that are hard to find on the market.

Most respondents, however, belong to the Engine room. This group, 56% of the sample, has less rare skills and is still behind on AI tools: only 11% use generative AI every day. The most critical point is that within this group just two people in five have access to the training they would need. Little surprise, then, that only half of these workers feel secure in their job, and that just 33% trust company leadership, against 75% among the pioneers.

Less and less training

The picture PwC paints is anything but encouraging, and it risks widening the gap between highly qualified workers and “ordinary” ones.

Artificial intelligence is set to spread further. Some 64% of respondents used it at work over the past year, 10 percentage points more than in 2025; the share of those using generative AI every day climbed from 14% to 22%, and 59% of the sample expects to use it more and more in the future.

In the meantime, though, in just twelve months the share of workers who say they have access to training and upskilling resources has fallen from 59% to 51%. And fewer than 49% of respondents know which skills they will need in the near future.

It is a trend that could turn into an employment and social problem very quickly.


Editor’s note

This article was originally published in Italian on money.it by Andrea Fabbri on October 16, 2026 as «L’AI crea lavoratori di serie A e di serie B. Cosa rivela il sondaggio PwC su 49.364 dipendenti». It has been translated and adapted for an international audience by the Money.it International desk.