Gartner forecasts $2.52T global AI spending boom in 2026

If Gartner’s $2.52 trillion figure is even directionally right, AI is no longer a ‘tech subsector’—it’s becoming a primary layer of global IT and capex. The report underscores that most of this spending is shifting into infrastructure: AI-optimized servers, accelerators, and cloud platforms that are effectively the substrate on which frontier models run. That’s the same hardware substrate AGI research will depend on.

The projection that AI infrastructure will reach roughly $1.37 trillion next year, driven by a 49% jump in AI-optimized servers, suggests that the NVIDIA-centered supply chain is still capacity-constrained and richly funded. That dynamic advantages players with preferential access to GPUs or equivalent accelerators: OpenAI–Microsoft, Google, Meta, Amazon, and a handful of well-capitalized startups. Smaller labs will increasingly have to differentiate with algorithmic efficiency, clever distillation, or novel architectures rather than brute-force scale.

Strategically, the ‘trough of disillusionment’ framing matters. Gartner sees enterprises exiting hype and consolidating on proven vendors, which could entrench incumbents and make it harder for new labs to gain commercial traction even if they have technically superior models. For the AGI race, that translates into a heavier bias toward the organizations that can turn research breakthroughs into production revenue fast enough to justify ever-larger infrastructure bets.