McKinsey's QuantumBlack consulting arm published new research this summer examining how corporate technology leaders are struggling to manage runaway artificial intelligence spending as adoption moves from pilot projects to company wide deployment. The analysis draws on a survey of 120 enterprise participants across five major industries conducted in May 2026.

The survey found that AI spending increases nearly fourfold as organizations move from isolated use cases into active, enterprise wide deployment. While 62 percent of organizations surveyed have already made that shift, 93 percent reported exceeding their AI budgets, and most expect spending to climb by at least 25 percent over the next 12 months.

Much of the overspending stems from limited visibility. McKinsey estimated that 20 to 30 percent of enterprise AI spend goes untracked because purchases are scattered across cloud providers, software vendors, and individual business units rather than managed centrally. Only 20 to 25 percent of companies surveyed have mature financial management practices in place for AI specifically.

The cost pressure mirrors a broader spending surge among the technology companies building AI infrastructure. Combined capital expenditures at Microsoft, Amazon, Alphabet, Meta, and Oracle climbed from 162.3 billion dollars in 2022 to 448.3 billion dollars in 2025, according to separate data from Epoch AI. By the fourth quarter of 2025, the five companies were spending a combined 140.6 billion dollars in three months alone.

McKinsey researchers said companies that actively manage AI consumption, including techniques like reusing stored prompts to cut repeated computing costs, have captured savings of 20 to 30 percent without reducing output.

Source: McKinsey & Company - https://mckinsey.com/capabilities/quantumblack/our-insights/the-cost-of-intelligence-how-cios-can-manage-ai-demand-at-scale