Does AI Destroy Jobs? A Myth Is Crumbling

The current discussion in Switzerland has been sparked by the new AI report from jobs.ch. It shows that the number of job postings for entry-level positions has declined by 32 percent since the end of 2022, with particularly sharp drops in AI-exposed sectors such as administration, marketing, finance, and IT. However, the study presents the picture in a somewhat more nuanced way, and the authors themselves acknowledge that, in addition to AI, economic factors also play a role. Media coverage, however, quickly jumps to the seemingly obvious conclusion that AI is systematically pulling the rug out from under the younger generation.

This is exactly where it’s worth taking a closer look. At the end of June, Ramp and Revelio Labs published one of the most comprehensive analyses on the topic to date. They linked the actual AI spending of over 21,000 U.S. companies to their employment data. The results directly contradict the conventional wisdom. Companies that use AI intensively and systematically increase their workforce by about 10 percent in the two years following implementation. For entry-level positions, the increase is even higher, at around 12 percent. The effect extends broadly across the organization, from engineering to sales to customer service. And it does not occur immediately, but only after six to twelve months, once companies have adapted their processes and completed the organizational learning phase.

This suddenly paints a different picture. In companies that use AI most intensively, AI isn’t destroying entry-level jobs—it’s creating additional ones. Anyone who simply signs up for a ChatGPT subscription won’t see this effect. On the other hand, companies that deeply integrate AI into their processes, products, and business models grow faster, become more productive, and hire additional people—including those just starting their careers.

These are important signals for Switzerland. First, we should take the drama out of the debate. The current decline in entry-level positions is real, but can largely be explained by economic conditions. Geopolitical uncertainty, protectionist measures in many countries, and high domestic costs are leading to general reluctance to expand the workforce. Therefore, singling out artificial intelligence as the main culprit is convenient, but—as is now becoming clear—not supported by empirical evidence. Second, we should ask the real question: Why hasn’t the Swiss economy made more progress in the productive use of this technology? Those who fall behind in AI adoption don’t just lose productivity; they also lose the growth momentum that is creating additional jobs elsewhere—especially for young people.

This leads to an economic policy agenda that is fundamentally constructive. Strengthen the framework conditions for investment in digitalization and AI. Expand continuing education without making it overly bureaucratic. Design regulations in a way that builds trust but does not stifle innovation.

“FUD”—Fear, Uncertainty, and Doubt, to throw a buzzword into the mix—is a poor guide for an aging, export-oriented economy facing a shortage of skilled workers. What Switzerland needs is less doomsday rhetoric and more sober analysis. The “staircase” that people warn so emphatically about does not arise where companies use too much AI. It looms where too little of it is actually put into practice.

This article was first published on July 16, 2026, by CH Media.