AI Productivity Rises While Reported EBIT Impact Stays Flat - Quasa.io

TEXT START: McKinsey’s 2026 Global Survey found that 80% of respondents who use AI in their work believed it had improved their individual productivity, while 37% attributed at least some positive EBIT contribution to AI—essentially the same share as in 2025.

The Dissection

The text is performing competent ROI hygiene. It separates personal productivity, workflow throughput, functional economics and enterprise EBIT, correctly exposing how AI enthusiasm inflates task-level gains into claims of corporate transformation.

Its deeper move is managerial containment: it reframes disappointing financial results as a workflow-design failure. The proposed remedy is redesign, measurement, ownership and agent deployment. That diagnosis is locally valid—and strategically incomplete.

The Core Fallacy

The article treats enterprise EBIT as the decisive endpoint and assumes that better workflow execution can convert AI-generated capacity into durable organizational profit.

Under the Discontinuity Thesis, that is not the terminal mechanism. P1 makes cognitive output cheaper and more abundant. P2 prevents firms from preserving stable human-only economic domains. Competition then pushes efficiency gains into lower prices, higher service expectations, excess capacity or greater output—not necessarily durable EBIT. The firm may become more capable while labor becomes less necessary.

The article mistakes a structural transition for an operations problem. Even the “high performers” it cites may simply be early Sovereigns capturing the advantage before it is competed away. Their redesign does not save the post-WWII system; it accelerates the removal of human labor from the productive circuit.

Hidden Assumptions

• AI-created capacity will find profitable demand rather than produce overcapacity or price compression.
• Firms can retain productivity gains instead of passing them to customers, platforms or competitors.
• Redeployment will remain available as roles are decomposed and automated.
• Better measurement and workflow ownership can preserve broad productive participation.
• Headcount reductions are the main signal of disruption, when wage erosion and declining labor necessity may arrive first.
• Enterprise EBIT captures the relevant social outcome. It does not capture the collapse of the wage-to-consumption circuit.
• The main obstacle is organizational incompetence rather than the competitive inevitability of cognitive substitution.
• Survey-attributed EBIT is meaningful enough to guide strategy, despite being self-reported and non-causal.

Social Function

Partial truth functioning as transition management and elite self-exoneration.

The article is too empirically disciplined to be simple copium. Its evidence ladder is sound. But by locating the failure in queues, governance and workflow design, it lets executives believe that competent management can preserve the old economic order. It converts an approaching labor-displacement crisis into a consulting backlog.

The Verdict

The article correctly shows that AI productivity is not automatically AI profit. It fails to confront the more terminal fact: even when productivity is successfully converted into profit, the result can be less human labor, weaker wages and a faster break in the mass employment → wage → consumption circuit.

Flat EBIT is not evidence that the system is safe. It is evidence that the transition has not yet fully reached the enterprise income statement. The managerial fixes described here may improve individual firms while making the social system more obsolete.