Image adapted from Getty Images
Artificial intelligence has rapidly evolved from a technological promise into one of the biggest investment races in corporate history. In 2026, AI-related investment could reach around US$800 billion, driven primarily by the construction of data centres, the purchase of chips, energy requirements and the wider infrastructure needed to support new AI models.
Microsoft, Amazon, Alphabet and Meta are among the companies investing most heavily in this expansion. At the same time, Nvidia, one of the biggest beneficiaries of the AI boom, is beginning to play an even greater role within the ecosystem. According to the Financial Times, the company is working with Wall Street groups on a structure capable of mobilising up to US$500 billion to finance artificial intelligence infrastructure.
The race continues. But the question being asked by the market is beginning to change: after so much money has been invested, where is the return?
Over the past few years, being left behind in artificial intelligence appeared to pose a greater risk than investing too much. Companies rushed to develop tools, hire specialists, integrate solutions and experiment with new forms of automation.
Now, a different phase is beginning.
Although there are clear examples of success, many companies are still struggling to translate their AI investments into significant cost reductions, productivity gains or new revenue. The debate has already reached investors, who are beginning to scrutinise the gap between rising expenditure and the returns actually being generated.
This does not mean that artificial intelligence has failed. Quite the opposite. Demand for computing capacity remains high, and the AI boom is driving activity across sectors far beyond technology, including energy, construction, cooling systems, semiconductors and financial markets.
What has changed is the expectation.
Investing is no longer enough. Companies need to demonstrate results. Artificial intelligence may now be entering its most important phase. The first was defined by discovery: everyone needed to experiment. The second was about investment: everyone needed to build. Now comes a third stage, one that is far more business-driven: everyone will need to justify.
For leaders and business owners, there is an important lesson in this shift. Adopting AI simply because competitors are doing the same is not a strategy. The technology needs to solve a specific problem, increase productivity, reduce costs or create new revenue opportunities.
After billions have been invested, the market is beginning to distinguish between those who have genuinely found a strategic application for artificial intelligence and those who have simply followed a trend.
And perhaps it is precisely this next phase that will determine the true winners of the AI race.
Can your company already measure the return on its investments in artificial intelligence — or is it still investing simply to avoid being left behind?
