AI investing: growth engine or a bubble waiting to burst?

Imagine an American economy standing not on consumption or productivity, but on the billions being poured into artificial intelligence.

According to several economists, without the tech giants’ massive AI investments, the United States would already have slipped into recession. Over the past two years, these companies have flooded the sector with capital, to the point where global AI spending could reach nearly 500 billion dollars by 2026.

Yet behind this frenzy of innovation, warning signs are piling up. The IMF estimates that around 60% of jobs in advanced economies are now exposed to automation.

Even OpenAI, the company behind ChatGPT, has reportedly suffered nearly 12 billion dollars in losses in a single quarter — weighed down by astronomical operating costs, from research to data centres and infrastructure.

So why do the GAFA (Google, Apple, Meta, Amazon, Microsoft) continue to pour money into this technological black hole?
Because they are locked in an AI arms race, where whoever reaches the next breakthrough first could secure a massive strategic advantage.

And in this race, none of them intends to finish last.

AI as the artificial engine of US growth

Artificial intelligence has become an unexpected pillar of recent US economic growth. A Deutsche Bank analysis even suggests that without AI-related investment, the US economy would have tipped into recession. Big Tech has rapidly expanded spending to develop new AI models, and the “Magnificent Seven” (Apple, Microsoft, Google, Meta, Amazon, Tesla, Nvidia) now account for over a third of the S&P 500.

In short, AI has become the temporary engine holding up the US economy, keeping growth alive despite the challenging environment.

Massive investment in infrastructure and software is boosting activity, but the foundations remain fragile: profits have yet to catch up.

For now, the economy is standing thanks to growth artificially fuelled by Big Tech’s spending.

The GAFA battle for AI dominance

While these colossal investments support economic momentum, their long-term profitability is increasingly questioned. Many AI initiatives currently cost far more than they earn. The case of OpenAI is telling: despite the popularity of ChatGPT, the company is said to have posted nearly 12 billion dollars in losses over the most recent quarter, for around 4.3 billion dollars in revenue across six months.

Generative AI models require extremely expensive infrastructure (high-end chips, enormous compute clusters) and constant research, placing heavy pressure on finances.

📊 Estimated AI investment table for 2025

AI is no longer a simple gadget: it has become the temporary engine of the US economy, powered by massive spending on servers, software and data centres. But the foundations remain fragile — profits are not keeping up.

The AI bubble and echoes of the 2000s dot-com era

At the end of the 1990s, exuberance around internet start-ups fuelled a speculative bubble. Young tech companies with no viable business model saw their valuations skyrocket, until the illusion burst in 2000. The dot-com crash led to bankruptcies across the sector and a major market correction. That episode remains a warning: technological enthusiasm can quickly turn into a financial mirage when profitability fails to materialise.

Jobs at risk: automation waiting in the wings

The AI boom is also reshaping the labour market. According to the IMF, nearly 60% of jobs in advanced economies are exposed to automation.

This time, it’s not only manual roles under threat.
Many highly skilled professions could be directly affected by the rapid rise of AI.

Of course, AI will also create new jobs: model trainers, data engineers, virtual agent designers…

But the transition will be harsh. Many workers will have to retrain quickly, often without enough support.

The challenge is clear: while AI boosts productivity, its benefits must also reach employees, not only shareholders.

How to invest smartly in AI

AI is attracting vast amounts of capital, but investing without strategy can quickly turn into speculation.

Here are a few principles to ride the trend without getting caught in the bubble

• ✅ Diversify: don’t put everything on Nvidia or OpenAI. Add related sectors such as energy, semiconductors or cybersecurity.
• ✅ Favour thematic ETFs: they reduce individual risk and capture the wider trend.
• ✅ Invest progressively (DCA): spreading purchases over several months helps reduce bubble effects.
• ✅ Check fundamentals: an “AI-labelled” stock without real profits is still a promise, not a safe bet.

📈 A few popular ETFs in the AI sector

* The TER (Total Expense Ratio) represents the annual management cost of an ETF.