Why AI is now a capital story
The current AI boom is often described in the language of models, products, and consumer adoption. That description is too light. The deeper reality is financial and physical. Artificial intelligence in 2026 is being built through one of the largest concentrated infrastructure spending waves in modern corporate history. Reuters reported in February that Alphabet, Amazon, Meta, and Microsoft are expected to spend about $650 billion in 2026 on AI-related infrastructure, up sharply from roughly $410 billion in 2025. On March 10 Reuters also reported that Citigroup raised its AI capital expenditure and revenue forecasts for 2026 to 2030, lifting its global AI revenue forecast to $3.3 trillion from $2.8 trillion. These numbers point to something larger than a software trend. They point to a financial order being reorganized around the expectation that compute-heavy AI systems will sit at the center of future growth.
This is why it is misleading to discuss AI as if the main issue were model cleverness. Clever models matter, but they do not scale without data centers, semiconductors, networking gear, power contracts, cooling systems, financing structures, and large anchor customers willing to make recurring commitments. The AI boom is therefore inseparable from a capital-allocation boom. Banks, equity markets, sovereign investors, utilities, chipmakers, cloud vendors, and specialized infrastructure companies are all being pulled into the same orbit. In effect, the industry is constructing a new investment thesis for the entire digital economy: intelligence capacity will justify enormous front-loaded spending today because it will become indispensable tomorrow.
Premium Gaming TV65-Inch OLED Gaming PickLG 65-Inch Class OLED evo AI 4K C5 Series Smart TV (OLED65C5PUA, 2025)
LG 65-Inch Class OLED evo AI 4K C5 Series Smart TV (OLED65C5PUA, 2025)
A premium gaming-and-entertainment TV option for console pages, living-room gaming roundups, and OLED recommendation articles.
- 65-inch 4K OLED display
- Up to 144Hz refresh support
- Dolby Vision and Dolby Atmos
- Four HDMI 2.1 inputs
- G-Sync, FreeSync, and VRR support
Why it stands out
- Great gaming feature set
- Strong OLED picture quality
- Works well in premium console or PC-over-TV setups
Things to know
- Premium purchase
- Large-screen price moves often
That thesis has persuasive elements. Enterprise adoption is clearly broadening. Reuters on March 10 cited Citi's view that enterprise demand was accelerating rapidly. OpenAI, Anthropic, Google, Microsoft, Meta, and Amazon all continue to push deeper into business and institutional use cases. Thomson Reuters' own 2026 professional-services report also pointed to a tipping point in adoption across legal, tax, accounting, risk, fraud, and government functions. When that evidence is paired with public-sector uptake, consumer normalization, and the desire of firms not to be left behind, it becomes easier to see why investors and executives keep raising spending plans rather than cutting them.
Yet there is a second side to the story. Reuters Breakingviews argued on March 11 that if OpenAI or Anthropic were to fail, the resulting shock could destabilize the broader AI boom because so much spending has been justified by the expected success of a relatively small number of frontier model companies. That is the most revealing vulnerability in the current financial order. Vast infrastructure commitments are being made on the assumption that demand will continue compounding, model firms will keep improving, customers will move from pilots to dependence, and monetization will eventually catch up with spending. If those assumptions weaken, the consequences would ripple outward beyond software valuations into chip demand, cloud revenue, credit exposure, data-center occupancy, and even regional energy plans.
Stacked confidence and stacked vulnerability
This is what makes the AI financial order different from an ordinary product cycle. Many layers of the economy are now being synchronized around the same expectation. Chipmakers expand because cloud companies build. Cloud companies build because model providers need more compute and enterprise customers may demand more inference. Utilities and governments adapt because data centers need power and transmission. Consultants and service firms invest because clients want implementation help. Public markets reward the cycle because growth expectations rise with each new spending forecast. The result is a stacked system in which confidence at one layer supports confidence at the next.
That can produce extraordinary momentum. It can also produce fragility. The more every layer depends on high utilization of future AI capacity, the more painful any demand shortfall becomes. This does not mean a collapse is inevitable. It means the boom is increasingly systemic. A frontier lab does not need to disappear entirely to cause stress; a slowdown in enterprise conversion, a plateau in model differentiation, regulatory friction, or a broad reassessment of pricing power could all matter. The rhetoric around AI often alternates between utopian inevitability and total bust. The truth is more structural. The sector is creating real capabilities and real demand, but it is also financing ahead of certainty on a grand scale.
One of the clearest signs of this structural turn is the rise of the so-called neocloud and AI infrastructure intermediaries. Reuters reported in March that French and British infrastructure plays such as Nebius and Nvidia-backed Nscale are raising large sums and building data-center capacity to serve the AI ecosystem. These firms are not household AI brands. Yet they matter because they translate capital markets into actual compute availability. Their role resembles that of rail builders, telecom carriers, or cloud wholesalers in earlier eras: they do not define the whole system, but without them much of the system cannot scale. AI therefore creates opportunities not only for the laboratories that capture headlines, but also for the less glamorous firms that turn financing into physical capacity.
Energy is now tied directly into this financial architecture. France is pitching nuclear-backed AI data centers. Germany is pushing domestically run compute. The United States has said it is ready to work with countries on civil nuclear development in light of data-center power demand. As data-center investment expands, utilities, grid operators, power developers, and industrial policy planners become part of the AI balance sheet whether they use that language or not. A model may look weightless on screen, but the revenue model behind it is being underwritten by assets that are anything but virtual.
Infrastructure firms, power demand, and narrative pressure
This is also why AI has begun to alter the hierarchy of strategic corporate relationships. In an earlier software cycle, application vendors often sat at the visible edge while underlying infrastructure remained abstract to most users. In the present cycle, infrastructure itself becomes a public story. Investors track GPU supply, hyperscaler capex, power availability, fiber routes, liquid cooling, and sovereign data-center deals because these have become direct constraints on the growth narrative. The capital story is no longer hidden underneath the product story. It is the product story's enabling condition.
There is another implication. The more money is poured into AI infrastructure, the greater the pressure on model providers and platform companies to show that this spending is not merely defensive. If firms are investing hundreds of billions collectively, they need customers, legislators, and markets to believe that real economic transformation is underway. This can intensify the social pressure to adopt AI before many institutions are fully ready. Nobody wants to look late to a general-purpose technology, especially when competitors are loudly declaring strategic urgency. As a result, the financial order can accelerate adoption not only through supply but through narrative compulsion. Spending itself becomes a signal that everyone else must move.
The user-facing result is familiar: more AI inside office suites, search engines, messaging platforms, development tools, shopping interfaces, and public-sector workflows. But the hidden driver is capital discipline of a peculiar kind. Firms that have committed huge sums must keep finding routes to embed AI deeply enough that the revenue logic remains plausible. This is one reason enterprise agents, government contracts, and sovereign AI partnerships matter so much. They are not side experiments. They are potential answers to the question of how to justify the scale of infrastructure being built.
The current order may still succeed spectacularly. It is entirely possible that AI-driven productivity, automation, and new service layers will be large enough to absorb the spending wave and make today's forecasts look conservative. But even in that case, the shape of the economy is changing. Intelligence capacity is becoming a strategic asset class. Compute is becoming a site of geopolitical and financial competition. The boundary between software company and infrastructure company is blurring. And a growing share of the future is being collateralized against the belief that machine-mediated reasoning will become ordinary across work, governance, education, and commerce.
The financial order behind the AI age
This is why analysts increasingly watch the financial plumbing of AI rather than only product launches. Debt markets, private-equity rounds, sovereign investment vehicles, utility agreements, and data-center preleases are all early indicators of whether the cycle is deepening or becoming overextended. In previous technology waves, capacity sometimes outran demand and then corrected sharply. In AI, the difference is that the capacity being built is so capital-intensive and so geopolitically entangled that any correction would not remain confined to a few public software names.
The political consequence is that AI finance is becoming public policy whether politicians admit it or not. When governments court data centers, revise power planning, offer tax incentives, or partner with frontier labs for defense and public services, they are helping decide who bears risk and who captures upside in the AI buildout. The financial order behind AI is therefore not just a market phenomenon. It is increasingly a public decision about what kinds of infrastructure, dependence, and concentration societies are willing to underwrite.
That is why the $650 billion figure matters symbolically as well as practically. It marks the point at which AI ceases to be a story about cleverness alone and becomes a story about order: who can finance the stack, who can host it, who can afford dependence on it, and who can survive if the cycle disappoints. The AI age will not be decided only by benchmark scores or viral apps. It will be decided by whether the vast system of capital, infrastructure, and institutional adoption now gathering around frontier models can sustain itself long enough to become the new normal.
Continue with OpenAI, Countries, and the Bid to Become National AI Infrastructure 🌐🏛️⚙️, China, Europe, and the Race for Sovereign Compute 🌏⚡🏭, and OpenAI, States, and the Race to Become Public Infrastructure 🏛️🤖.
Books by Drew Higgins
Prophecy and Its Meaning for Today
New Testament Prophecies and Their Meaning for Today
A focused study of New Testament prophecy and why it still matters for believers now.