Demand for the complex IT infrastructure required to develop and operate artificial intelligence—databases, server farms, high-speed networks, power delivery and cooling—has grown significantly in recent months. That rising demand has increased investor interest in semiconductor manufacturers and related suppliers, although equities in those companies have experienced substantial corrections in recent weeks as investors took profits.
The market moves may reflect more than short-term profit-taking: the breakneck expansion of the AI sector and the euphoria around the technology appear to be cooling somewhat. Nonetheless, investments in AI infrastructure are expected to remain very large over the coming decade; projections cited in the reporting foresee that such spending through 2035 will represent one of the largest capital concentrations seen in modern technology history.
Corporate capital expenditure (CAPEX) focused on AI in such a compressed timeframe would be practically unprecedented. Industry attention has shifted from primarily developing AI models to building the physical infrastructure needed to run those models at scale. Current trends suggest AI-related investment could exceed $700 billion in 2026 and approach $3 trillion by 2028.
The beginning: ChatGPT and the market response
The rapid transformation traces back only a few years. In 2022, OpenAI introduced ChatGPT, which allowed users to interact with an AI in natural language and marked a turning point for widespread adoption.
Investment patterns in this cycle were atypical: instead of following classic venture-capital stages, large technology companies—most notably Microsoft—quickly saw strategic value and provided significant funding and partnership arrangements for equity stakes. Microsoft also supplied the hardware infrastructure necessary to operate the ChatGPT service, accelerating deployment.
OpenAI’s financial performance reinforced that investment case: the company posted exceptionally strong revenue growth by 2023, and available information indicates revenues have more than tripled year over year since then.
Why this matters
The shift in emphasis from model development to infrastructure deployment affects more than technology: it reshapes corporate investment priorities and capital allocation across markets. The scale and speed of AI-related hardware and infrastructure spending could materially impact the global semiconductor industry, data center expansion, and demand for energy and cooling solutions.
At the same time, recent stock-market corrections signal investors are rebalancing positions after profit-taking and reassessing risks. A reduction in euphoria can increase near-term volatility, yet if current investment forecasts hold, substantial long-term infrastructure spending will continue to concentrate capital in this sector.
Mentioned organisations
- OpenAI
- Microsoft
- Gránit Alapkezelő (appears as a referenced keyword in the source)
Monitoring developments will be important for both technology providers and investors, because capital-allocation decisions in the coming years are likely to shape the structure of the AI ecosystem and the competitiveness of associated industries.
