AI Monetization Is Reinforcing the Infrastructure Buildout

AI monetization is beginning to catch up with infrastructure costs, making the AI investment cycle increasingly self-reinforcing.

• Global X estimates that AI-linked revenue across 15 platforms and providers will rise from $146 billion in 2025 to $317 billion in 2026, $518 billion in 2027, and nearly $1.5 trillion by 2030, implying a 60% CAGR from 2025 to 2030.
• Frontier-lab and cloud monetization is accelerating: Anthropic’s annualized revenue surpassed an estimated $75 billion in July 2026, OpenAI’s exceeded $40 billion, AWS’s AI business passed $25 billion, and Microsoft reported more than 30 million paid Copilot seats.
• Contracted demand is financing further capacity, with Microsoft’s commercial remaining performance obligations up 84% year over year to $678 billion, Google Cloud’s backlog at $514 billion, and CoreWeave’s backlog above $125 billion.
• The infrastructure buildout remains exceptionally large: nearly $6.7 trillion could be invested in AI infrastructure between 2025 and 2030, while hyperscaler capex is projected at $880 billion in 2026 and roughly $1.13 trillion in 2027.
• Supply bottlenecks are emerging in enabling hardware: HBM spending could increase from $35 billion in 2025 to $100 billion in 2028 and $130 billion in 2030, while inference may represent nearly 60% of AI data-center demand by 2030 and agentic AI could drive a 24-fold increase in token consumption between 2026 and 2030.

Takeaway: Hardware and physical infrastructure should remain the clearest near-term beneficiaries, but falling inference costs and rising recurring workloads should progressively shift value toward model platforms, software, applications, and cybersecurity; the thesis depends on AI revenues continuing to scale alongside usage.