88% of Orgs Use AI, but Only 39% Can Attribute EBIT Impact

The uncomfortable truth in enterprise AI right now isn't model capability — it's accountability. McKinsey's State of AI 2025 found 88% of organizations use AI in at least one function, but only 39% can attribute any EBIT impact to it, exposing a measurement and ownership vacuum. If AI is 'doing the work, not just supporting it,' as Irina Shymko argues, then the technical problem facing AI leaders is not which model to deploy but how to govern usage, architecture, and cost at the resource level.

Irina Shymko's HackerNoon essay opens with a scene that has become familiar in enterprise boardrooms: a leadership team is asked how much it spends on AI and what outcomes that spending produces, and nobody in the room can answer. Her central argument is that organizations are still managing generative AI as if it were conventional software — a line item to be procured, licensed, and tracked — when it has in fact become an enterprise resource that is doing the work, not merely supporting it. That distinction reframes the core question from 'how much are we spending on AI?' to 'what outcomes is it producing, and who owns them?' — a question two decades of IT management tooling was never designed to answer.

“

The urgency is driven by scale: citing Gartner, the essay puts global AI spending at a projected $2.59 trillion in 2026, up 47% year over year.

Every major technology wave, Shymko notes, has forced the creation of a new management discipline. Cloud turned compute into an on-demand resource and gave rise to cloud economics and FinOps. Software-as-a-service produced sprawling subscription portfolios, which in turn created software procurement, access management, and vendor governance. Generative AI is the next wave in that sequence, and most companies have not yet built the discipline it demands. The urgency is driven by scale: citing Gartner, the essay puts global AI spending at a projected $2.59 trillion in 2026, up 47% year over year. At the same time, McKinsey's State of AI 2025 found that 88% of organizations now use AI in at least one function, yet only 39% can attribute any EBIT impact to it — and for most of those, the impact is under 5%.

That is an extraordinary gap: nearly nine in ten companies are using AI, but fewer than four in ten can connect it to profit, and most of those connections are marginal. Shymko's contention is that the enterprise conversation has been misfocused on cost rather than outcomes. There is legitimate cause for concern on cost: Gartner predicts that through 2028, at least half of GenAI projects will overrun their budgets because of poor architectural choices and a lack of operational know-how. But she argues cost is a symptom rather than the disease. The deeper problem is a lack of visibility into what companies are actually paying for — finance teams see AI spend scattered across model usage, infrastructure, and data, with no single owner and no clear line to outcomes.