Goldman: Four tech giants to spend more than Japan's GDP on AI by 2030

An investment scale without historical precedent

Goldman Sachs has published sharply revised estimates for AI infrastructure investment, painting a picture of an investment cycle the business world has never seen before. According to the bank's updated forecasts, as reported by ForexLive, the four major technology companies — Meta, Microsoft, Amazon, and Alphabet — will collectively invest $5.3 trillion in capital expenditures through the end of 2030.

To put that figure in perspective: it exceeds the gross domestic product of Japan, the United Kingdom, India, and France — each considered individually. When the broader industry is included — encompassing data centers, power infrastructure, and compute capacity — Goldman estimates total investment could reach $7.6 trillion within the same five-year window.

The acceleration is already underway

Perhaps the most immediately relevant figure for market participants is this year's expected capex level. Goldman estimates that the four companies will spend up to $725 billion in capital investments during 2025 — more than double the $360 billion they spent in 2024. The pace of growth indicates that the investment cycle remains in an expansion phase, nowhere near any natural plateau.

This means that suppliers of power, construction services, and industrial supply chains are already feeling the effects in their earnings cycles — not merely at some hypothetical 2030 horizon.

Private markets are absorbing a larger share

One structural observation Goldman highlights is that private markets will increasingly finance the build-out of AI infrastructure. Private data center construction has, according to the bank, accelerated significantly in recent years, and Goldman expects this share to grow further.

This is a signal that listed equity markets should take note of: a growing portion of the returns from AI infrastructure will flow to private capital vehicles rather than publicly listed companies. The hyperscalers thus appear to be moving faster than public capital markets can efficiently absorb.

The return question remains unanswered

Despite the impressive figures, investor concerns about returns have not disappeared. ForexLive notes that the gap between capital deployed and the revenues AI products have actually generated remains a live topic of debate in the markets.

Goldman addresses this implicitly by characterizing the current period as an early phase of a durable cycle rather than a speculative peak — but the bank does not quantify when or how returns will materialize. Critics will therefore continue to argue that the question of commercialization pace remains unanswered.

Implications for capital markets

For investors, the Goldman report signals that data center infrastructure, power supply, and industrial supply chains have multi-year tailwinds. The figures confirm that the AI investment cycle has years of runway remaining.

Direct exposure to this theme on the Oslo Stock Exchange is limited, but Norwegian industrial companies and power suppliers with exposure to European data center construction — a rapidly growing market — may benefit indirectly from the trend. Otherwise, the implications of Goldman's analysis apply primarily to global capital markets and sectors such as power, construction, and semiconductors.

The most important takeaway from Goldman is nonetheless straightforward: four companies have made a collective decision to invest at a scale that has no historical parallel in the business world.