Over the past week, global markets have been forced to confront a growing contradiction at the heart of the artificial intelligence boom. On one hand, Big Tech has doubled down on AI with unprecedented levels of capital spending. On the other, investors have begun to question whether the economic logic of this expansion is keeping pace with its ambition.
Together, these tensions have reshaped market sentiment, driven sharp sector rotations, and reignited long-standing fears of a technology investment bubble — this time built not on dot-coms, but on data centres, AI chips and computational scale.
A Spending Wave Without Precedent
At the centre of the week’s debate was the revelation that the world’s largest technology firms plan to spend an estimated $660 billion on capital expenditures in 2026, largely tied to artificial intelligence infrastructure. Amazon, Microsoft, Alphabet and Meta are leading the charge, committing vast sums to data centres, custom chips and cloud expansion.
The scale of this spending is historic. Compared with just two years ago, projected AI-related investment has more than doubled. For proponents, the logic is clear: AI represents a foundational technology, and control over compute capacity will define future competitive advantage.
However, markets reacted less enthusiastically. Despite solid earnings, several Big Tech stocks fell as investors questioned whether returns could justify such aggressive outlays in the near term.
From Enthusiasm to Scrutiny
This shift in tone marks a notable change from the optimism that defined much of 2024 and 2025. Then, AI exposure was broadly rewarded. This week, it became a source of concern.
Investors are no longer asking whether AI will transform the economy — a question largely settled — but who will capture its value, and when. The gap between capital intensity and visible revenue growth has become harder to ignore.
Nowhere was this more evident than in the software sector. Following the launch of new AI automation tools by Anthropic, global software and data analytics stocks sold off sharply. Markets interpreted the tools as potential substitutes for parts of the enterprise software stack, prompting a reassessment of long-term subscription models.
Software Feels the Pressure
The sell-off in software stocks highlighted a deeper anxiety: that AI may not simply enhance existing business models, but actively compress them.
Legal tech, professional services software and data analytics firms were among the hardest hit. These are areas where AI systems increasingly demonstrate the ability to automate structured, repeatable work — tasks that once justified high licensing fees.
For years, enterprise software benefited from predictability and lock-in. This week, that assumption was tested. Investors began pricing in a future where AI tools reduce dependency on traditional platforms, even if they do not replace them outright.
A Global, Not Local, Reckoning
Importantly, these developments were not confined to U.S. markets. European software firms tracked the decline, while Asian technology stocks followed suit. The reaction underscored how deeply interconnected global capital markets have become around the AI theme.
At the same time, not all technology assets moved in tandem. Semiconductor and hardware firms showed greater resilience, reflecting their role as infrastructure providers rather than application-layer competitors.
This divergence suggests markets are beginning to differentiate between AI as capacity and AI as product — a distinction that may shape investment strategies in the months ahead.
Bubble Fears, Reframed
Talk of an “AI bubble” has returned, but in a more nuanced form than past technology cycles. Unlike the dot-com era, today’s leading AI investors are profitable companies with strong balance sheets. The risk lies less in speculation than in overcapacity and mispriced expectations.
The question facing markets is not whether AI spending will continue — it almost certainly will — but whether returns will arrive fast enough to satisfy investors accustomed to rapid growth.
An Inflection Point for 2026
As the week closes, one conclusion stands out: AI has entered a more mature phase of its financial life cycle. Capital is no longer flowing on promise alone. Execution, efficiency and revenue clarity now matter more than scale for its own sake.
For policymakers, investors and technology leaders alike, this moment represents an inflection point. The coming months may determine whether AI’s current trajectory leads to sustainable transformation — or to a costly period of recalibration.
What is clear is that confidence, not capability, has become the most contested variable in the global AI economy.
Subscribe to EuroLuminant for independent European journalism.
