Goldman Sachs Forecasts $700B AI Capex Surge Amid Energy Supply Constraints

The artificial intelligence revolution is entering a capital-intensive second phase, shifting from software development and model training to a massive build-out of physical infrastructure. Goldman Sachs recently projected that AI-related capital expenditures (Capex) could reach a staggering $700 billion in 2026. This forecast underscores a pivot in the market: while the first wave of the AI trade was dominated by chipmakers like Nvidia, the next chapter belongs to the 'picks and shovels' of the physical world—specifically power generation and data center infrastructure. The sheer scale of this investment reflects a belief among hyperscalers that the competitive advantage in AI will be determined by who can secure the necessary hardware and, more importantly, the electricity to run it.

As this $700 billion flows into the market, energy availability has emerged as the most significant bottleneck. Modern AI data centers consume significantly more power than traditional facilities, requiring high-density cooling and a constant, reliable supply of electricity. For the world’s largest technology companies, many of which have aggressive net-zero carbon mandates, this creates a complex challenge. They cannot simply plug into any grid; they require massive amounts of carbon-free energy. This requirement has transformed the utility sector from a defensive, low-growth industry into a high-stakes partner for the tech elite. The demand for green power is now outstripping supply in many key data center hubs, leading to a race for long-term power purchase agreements (PPAs).

“

Goldman Sachs recently projected that AI-related capital expenditures (Capex) could reach a staggering $700 billion in 2026.

Brookfield Renewable (BEP) has emerged as a primary beneficiary of this trend. The company, which manages one of the world's largest platforms for renewable power and decarbonization solutions, is uniquely positioned to solve the energy bottleneck for AI giants. A landmark deal with Microsoft serves as the blueprint for this new era: Brookfield committed to delivering over 10.5 gigawatts of renewable energy capacity to Microsoft between 2026 and 2030. This agreement is nearly eight times larger than any previous corporate PPA, highlighting the unprecedented scale of power required by the next generation of AI workloads. With a global portfolio spanning hydro, wind, solar, and storage, Brookfield offers the geographic diversity and operational scale that hyperscalers need to expand globally.