Asger Risom
Despite widespread adoption, most companies report little real impact from AI—highlighting a growing gap between corporate hype and measurable productivity gains.
Artificial intelligence has spread rapidly across corporate America, dominating earnings calls, strategy decks and investor presentations. But when it comes to actual results, many companies are still waiting.
If it feels like AI should already be transforming productivity and jobs, new data suggests that expectation is running ahead of reality.
According to Fortune, nearly 90% of companies report that AI has had no meaningful impact on employment or productivity over the past three years.
Adoption without impact
On paper, AI is widely adopted.
A study of around 6,000 executives across the U.S., U.K., Germany and Australia found that roughly two-thirds of firms say they use AI. Yet in practice, that usage remains shallow, averaging just 1.5 hours per week.
A quarter of companies reported not using AI at all, highlighting a gap between corporate messaging and day-to-day operations.
A growing disconnect
The mismatch between hype and outcomes is becoming harder to ignore.
Executives continue to describe AI as transformative, and many expect it to boost productivity and output in the coming years. But those gains have yet to appear in broader economic data.
“AI is everywhere except in the incoming macroeconomic data,” Apollo chief economist Torsten Slok wrote, pointing to the absence of clear effects in employment, inflation or productivity.
When more becomes less
Some research suggests the way AI is being used may be part of the problem.
A study by Boston Consulting Group found that while limited use of AI tools can improve productivity, overuse has the opposite effect. Workers using multiple tools reported “brain fog” and more frequent mistakes.
At the same time, confidence in AI is declining even as adoption rises, suggesting growing frustration with its real-world usefulness.
Echoes of an old paradox
Economists see a familiar pattern.
In 1987, Nobel laureate Robert Solow famously noted that computers were visible everywhere except in productivity statistics—a contradiction that became known as the productivity paradox.
“You can see the computer age everywhere but in the productivity statistics,” he wrote.
Today, AI appears to be following a similar path: widely deployed, heavily promoted, but not yet reflected in measurable gains.
Waiting for the payoff
That doesn’t mean the impact won’t come.
The IT revolution also took years to translate into productivity growth, which only accelerated in the late 1990s after businesses adapted their processes.
Some economists believe AI may follow a similar “J-curve,” with slow early returns before a sharper payoff.
It’s not the tech—it’s the use
For now, the key issue may not be AI itself, but how companies implement it.
Many firms are still experimenting rather than fully integrating the technology into core workflows, limiting its impact.
“The value creation is not the product,” Slok said, “but how generative AI is used and implemented in different sectors in the economy.”
A gap between promise and reality
AI may be everywhere in business—but until it changes how work actually gets done, its economic impact will remain limited.
That gap between promise and reality is becoming one of the defining questions of the current tech cycle.
Sources: Fortune, NBER, MIT
