Artificial intelligence does not steal jobs… it quietly rewrites them

In the ongoing debate about artificial intelligence, there is a common tendency to overemphasize the idea of “job disappearance.” Yet what is happening on the ground is less dramatic—and at the same time more complex: jobs are not disappearing; they are being rewritten from within.

The real transformation is not about how many jobs will remain or vanish, but about how they are being broken down into smaller tasks, some of which are assigned to algorithms, while others are recombined into a narrower and more specialized human domain. This is precisely what the OECD highlights when it describes the shift from viewing the “job as the unit” to viewing “tasks as units that can be redistributed.”

But what appears clear in reports takes on a very different form when it collides with the real labor market.

In modern workplaces, it is no longer easy to define jobs in the traditional way. An accountant no longer simply performs calculations; a large part of the work is now tied to reviewing outputs generated by automated systems. A customer service agent is no longer always the first point of contact, but sometimes a second layer after intelligent systems handle the initial responses. Even data analysts no longer start from scratch, but from ready-made outputs generated by AI tools.

This change does not happen all at once; it gradually seeps into institutions until the distinction between a “full job” and a “bundle of tasks” becomes increasingly blurred.

In the Arab world, this process takes a more complex path. Historically, the labor market was not built on the idea of job decomposition, but on the job as a fixed unit: a clear title, a defined description, and a relatively linear career path. Therefore, the entry of AI does not simply mean upgrading tools—it forces a redefinition of how work itself is organized. This is what makes the transition less smooth than global reports might suggest.

The OECD, in its labor market analyses, points out that the main risk is not job loss, but a growing gap between the skills produced by education systems and those demanded by the economy. This gap is not new, but it has become more visible with the acceleration of digital transformation.

Some estimates show that a significant share of jobs in advanced economies now include tasks that can be partially or largely automated, meaning that change is no longer confined to low-skilled jobs, but extends to the middle-skilled workforce as well.

However, the real difference lies not in exposure to change, but in the ability to adapt and rapidly reshape skills.

In the Arab region, the issue is not limited to a skills gap. The education system still operates on a logic of stability: education is expected to last for many years of work. But this assumption is eroding quickly in an economy where tools evolve faster than graduation cycles.

What is becoming increasingly clear is that a large portion of today’s required skills did not even exist a few years ago, creating a persistent time lag between education and the labor market.

With the entry of AI into workplaces, jobs are not disappearing outright, but undergoing a quieter transformation: a gradual redistribution of tasks. Routine tasks are shrinking, while analytical and supervisory tasks are expanding. Meanwhile, the human role is being repositioned toward areas more closely linked to context and decision-making.

But this transformation does not affect everyone equally. Some jobs are being rapidly reshaped, while others are being gradually reduced without clear alternatives, creating an uneven redistribution within the labor market itself.

In the Arab context, this shift intersects with an economy that still relies, to varying degrees, on the public sector and traditional service activities. This makes AI less of a “technological revolution” and more of a test of institutional flexibility and the ability to reorganize.

In many cases, the issue is not access to technology, but the speed at which it can be absorbed into administrative and educational structures that were not designed for such rapid change.

Looking at the future through the OECD framework, three broad trajectories can be identified in the region.

There is a slow-adaptation path, where technology enters faster than institutions can adjust, widening the gap between skills and the labor market.

There is a consumption-driven path, where ready-made solutions are imported without building internal knowledge capacity.

And there is a third, more difficult but more sustainable path: re-linking education to the labor market so that skills become continuously evolving capabilities rather than one-time credentials.

In the end, artificial intelligence does not appear as an external force threatening the labor market, but as a factor that reveals its internal structure and reshapes it from within. It does not create imbalances, but accelerates and amplifies them.

The question that arises today is no longer about technology itself, but about how work is organized: how can a labor market continue to function when its rules are constantly changing?

In this context, the challenge is no longer entering the labor market, but remaining within it while it continuously redefines itself.