The New AI Boom Is Bigger, Heavier And More Vulnerable Than The Last One

The AI boom has not slowed in 2026; it has become more industrial, more expensive and more exposed to real-world constraints. Reuters reported at the end of March that Big Tech’s 2026 AI spending is facing an energy shock test, with estimated investment around 635 billion dollars now colliding with rising electricity prices and tighter power markets.​

That figure alone captures the scale of the transformation. AI is no longer mainly a software story. It is a story about data centers, chips, cooling systems, transmission capacity, real-estate procurement, talent, and long-term power contracts. The more the industry scales, the more it behaves like a utility-heavy industrial complex rather than a pure digital platform.

Reuters Breakingviews has already warned that the energy shock could derail the AI boom if companies are unable to preserve returns on the massive amounts of capital being deployed. The concern is not theoretical. Every increase in power cost can reduce the profitability of the next data-center tranche, which means AI economics may be more sensitive than many investors assumed when spending was still framed as “growth investment.”

What makes this cycle different from the last technology wave is that investors are becoming much less forgiving. Bond-market pressure, share-price volatility and growing skepticism about the path from AI capex to AI revenue are all forcing a sharper reassessment of business models. The market is no longer treating AI spending as automatically value-accretive. It now wants evidence.

The evidence problem extends to governance. Reuters has reported that AI accountability still lags behind the pace of deployment, and that boards need to push for better transparency, training-data controls and safeguards before AI becomes embedded in critical systems. In Asia, that caution is becoming law, with Vietnam and South Korea already enforcing new AI frameworks and other jurisdictions moving toward similarly strict rules.

The result is a technology sector that is still growing, but no longer on easy terms. Companies that provide the underlying infrastructure—energy management, cooling, networking, advanced semiconductors and enterprise-grade compliance—may benefit more than consumer-facing AI products that have yet to prove durable monetization.

There is also a strategic dimension. AI is increasingly tied to national industrial policy. Governments want domestic AI capability, domestic compute capacity and better control over critical digital infrastructure. That means the boom is becoming more capital-intensive and more geopolitically sensitive at the same time.

For businesses, the lesson is simple: the future of AI will be decided as much by power grids and boardrooms as by model architecture. The companies that win in 2026 will be the ones that can combine technical innovation with energy discipline and regulatory maturity.