AI Destroys Jobs, But Boosts Stocks: How Long Will This Last?

AI, an Ultra-Bullish Driver for Tech Stocks

For the past two years, artificial intelligence has become the main fuel for stock markets. American tech giants are showing staggering valuations, driven by the promise of multiplied productivity. Microsoft is now valued at $3.71 trillion, Alphabet at $4.33 trillion, and Amazon at $2.96 trillion.

The logic is simple. A company that replaces 10,000 employees with AI agents sees its costs drop, its margins swell, and its profits explode. On paper, the equation is perfect for shareholders. As analyst Firisis summarizes, job destruction is currently "bullish" because it reflects the true extent of AI adoption in the economy.

This first phase, characterized by productivity gains that boost margins without massively destroying the employment fabric, resembles a golden age for investors exposed to tech. Companies that automate gain competitiveness, while those selling AI see their revenues soar.

The Paradox That Could Flip the Market

The reasoning becomes complicated as we broaden the focus. The employees eliminated on one side are also consumers on the other. Fewer jobs mean mechanically less income, less purchasing power, and therefore less final demand.

What is optimal for each company taken in isolation can become toxic when all companies apply the same strategy simultaneously. An ultra-productive economy can produce gigantic volumes of goods and services, but if the gains are concentrated among capital holders, who buys all this production?

Firisis describes a three-phase sequence. The first phase where AI increases productivity without destroying jobs, very favorable to markets. The second phase where agents actually replace human labor, exploding the profits of AI winners. Then a third phase, the tipping point, where the cumulative decline in income finally weighs on demand and thus on the companies themselves.

Taxing Robots, An Idea That Resurfaces

It is precisely at this stage that an old proposal resurfaces: taxing robots. The idea is not new. In 2017, socialist candidate Benoît Hamon made it one of the flagship measures of his presidential campaign, proposing to levy a tax on "the wealth created by robots" to fund a universal basic income.

Nobel Prize-winning economist Robert Shiller has also advocated for a moderate tax on robots, conceived as a transitional tool to finance the retraining of displaced workers.

In the United States, the proposal is back on the table. Democratic representative Greg Casar recently introduced a bill aimed at taxing AI companies to fund job creation. Elon Musk, for his part, regularly mentions the hypothesis of a universal income financed by the productivity gains of AI.

The Legal Obstacles of a Robot Tax

But the idea faces several obstacles. A study published in 2025 by jurists from the IBFD highlights that directly taxing a robot implies recognizing it as a legal entity, which does not currently exist in either French or European law.

Thus, it will be necessary to monitor the moment when job destruction accelerates faster than productivity gains. Until then, the debate on the redistribution of AI gains, robot tax or universal income, is just beginning.