Billions Spent, Little to Show: The AI Investment Boom Hasn't Boosted Productivity

In 1987, economist Robert Solow famously noted that computers were visible everywhere except in the productivity data. As 2026 unfolds, a new version of that paradox is emerging. Despite historic corporate spending on artificial intelligence, the U.S. economy has yet to see a significant boost in output, leaving many to wonder if we’re repeating history.

A recent survey of Fortune 500 CEOs, reported by Fortune, finds a stark gap between enthusiasm and results. While investment in AI platforms and tools is massive, most executives admit they cannot yet point to measurable, company-wide productivity gains. The technology is here, they say, but weaving it into the fabric of their organizations is proving harder and slower than expected.

The scale of spending is unprecedented. Tech giants have committed over $200 billion recently to AI infrastructure, with total global investment projected to near $1 trillion. Yet national productivity growth remains modest. This disconnect mirrors the late 20th century, when huge investments in information technology took nearly a decade to show up in economic statistics.

CEOs cite familiar hurdles: pilot projects that don’t scale, poor data quality, employee resistance, and the complexity of fitting AI into old systems. Economists like MIT’s Erik Brynjolfsson describe a “productivity J-curve,” where output initially dips as companies learn and reorganize before potentially soaring. True gains, he argues, won’t come from simply buying AI, but from redesigning entire business processes around it.

Investors are growing impatient. After a rally in AI stocks, they now press executives for evidence of returns. The parallel to the dot-com era is clear—massive infrastructure spending preceded a bust, but later enabled a new digital economy. The question is how long the wait will be.

Some factors could speed things up: AI tools are becoming more capable and easier to use. But the organizational and workforce challenges are profound. AI doesn’t just augment work; it can replace it, creating anxiety that slows adoption.

The lesson from Solow’s paradox is that technology alone isn’t enough. Productivity followed when businesses reinvented themselves around computers. For today’s CEOs, the message is similar. The hard work of transformation, not the check written for the technology, will determine who eventually profits from the AI age.