Skip to content
Advertisement
80% of workers say AI makes them more productive, but only 37% of companies see profit

80% of workers say AI makes them more productive, but only 37% of companies see profit

Google / Wikimedia Commons (Public domain)

McKinsey's latest AI survey finds the gap between personal efficiency and bottom-line results hasn't budged in a year

Workers love their AI tools. The finance department is still waiting for the invoice to pay off.

A McKinsey survey found that 80% of employees using AI say it makes them more productive. Yet only 37% of organizations can link AI to any positive impact on earnings before interest and taxes, or EBIT. That number is unchanged from last year.

What McKinsey actually found

The findings come from McKinsey’s report, “The State of AI in 2026: On the Road to ROI.” The consultancy surveyed 1,719 executives across 97 countries between May 4 and June 8, 2026.

Use of AI in three or more business functions rose to 56%, up from 51% a year earlier. Companies reporting enterprise-wide AI implementation climbed to 44%, from 38% the previous year.

Only 6% of organizations qualified as “AI high performers” in McKinsey’s framework. To earn that label, a company had to attribute at least 5% of its EBIT to AI and report substantial impact.

Advertisement

What separates the winners

The high performers did not simply buy better software. According to McKinsey, they rebuilt how work gets done around the technology.

Nearly 75% of high performers reported substantial business transformation tied to AI. Among lower-performing organizations, that figure was about 25%.

High performers showed 3.3 times more interest in extensive AI-driven business transformation over the next three years.

The report’s core argument is that individual efficiency does not automatically flow to the bottom line. McKinsey points to structural changes such as headcount adjustments or shifting people toward higher-value work. Without those moves, the hour saved by drafting an email with a chatbot quietly evaporates into another meeting.

Why the paradox persists

The disconnect has a name in research circles: the “AI productivity paradox.” Rapid adoption, in this view, does not produce proportional financial returns on its own.

What stands out in McKinsey’s data is the lack of movement. A full year of broader rollouts produced no change in the share of companies seeing EBIT impact.

What this means for companies and investors

For executives, the message is uncomfortable but clear. Rolling out AI widely is not the same as getting value from it, and the survey shows a year of scaling did not fix that.

Companies that treat AI as an add-on to existing processes appear likely to stay in the 63% that see no earnings impact. The ones willing to redraw org charts and workflows are the ones showing up in the 6%.

There is a workforce dimension as well. McKinsey lists headcount adjustments among the structural changes that turn individual efficiency into corporate profit. The alternative path, moving people into higher-value work, requires management to define what that work actually is.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.