Automatic translation into Spanish is available for this page. Translate this page
A blue, tiered pyramid graphic, featuring people of all kinds, as well as robots in one section at the base, zoomed in on with a magnifying glass

OPINION PIECE - The real danger of AI is not just unemployment, it’s inequality

Esta página también está disponible en

Pilvi Torsti, Director of the European Training Foundation (ETF)

Artificial intelligence is often framed as a looming threat. The dominant theme is, “will AI take our jobs?”

But this is the wrong question. Recent analysis by the Turin-based European Training Foundation (ETF), drawing on global evidence, suggests something more subtle, and potentially more dangerous.

For now, AI is not (yet) destroying jobs at scale. Even in advanced economies, only a small share of roles currently face a high risk of full automation. Instead, AI is reshaping how work is organised, who benefits, and who is left behind.

As asserted by Alexandra Roulet in an opinion piece published in Libération on 7 April, AI optimists insist on its capacity to accelerate worker productivity. For some, this is already the case. Across sectors and countries, however, AI is also redistributing opportunity. 

For many lower-skilled and older workers, jobs are more intense, fragmented and controlled. Algorithmic management systems track performance in real time, standardise tasks, and reduce the scope for discretion or creativity. The gap between “augmented” jobs and “managed” ones continues to widen.

In parallel, AI is reshaping career pathways. Entry-level roles are particularly exposed. Tasks once performed by junior staff are increasingly handled by AI or consolidated into fewer roles.

For young people entering work, this raises a critical question: how will they gain the experience needed to progress?

These shifts reinforce existing inequalities between workers, and increasingly, between countries. This too should be a concern for Europe.

The EU’s economic, social and political future is intertwined with neighbouring economies. Many of these have entered the AI era with weaker digital infrastructure, lower levels of skills development, and limited regulatory capacity.

As AI adoption accelerates, this inequity will only intensify. Without effective governance, AI can amplify existing vulnerabilities, entrenching low-wage, low-security work in global digital supply chains.

It can deepen divides between those who can leverage such technology and those who cannot.

Inequalities can manifest in the form of instability, social tension and, ultimately, increased migratory pressures. In this context, the real challenge is how to ensure the benefits of AI are fairly distributed within societies and between them.

The impact of AI on labour markets is not dictated by technology alone – it is shaped by policy choices, institutional frameworks and investment decisions. The same tools that can intensify inequality can also broaden opportunity, if they are deployed with that aim in mind.

By investing in skills at home and abroad, Europe can support inclusive education and training systems that enable workers to adapt, rather than be left behind. By strengthening governance frameworks, we can better protect workers’ rights and ensure AI is used to “augment” human capabilities, not undermine them.

Supporting skills development beyond the EU is a strategic investment in our own stability, resilience and shared prosperity.

Artificial intelligence will continue to transform the world of work – greater inclusion or deeper division will depend on the choices we make now.

The question is no longer if AI will change work, but whether we are prepared to shape that change before inequality becomes entrenched.

This op-ed was originally published in French-language daily Libération on 21 May 2026.

Media relations