AI’s productivity boom will likely create a ‘winner-takes-all’ economy, top EY economist warns

AI was sold as the great equalizer. Gregory Daco, EY-Parthenon’s chief economist, thinks that pitch is backwards. His line is blunt: productivity growth protects margins, not income. In other words, the gains from better machines do not automatically wash through to workers; they can just as easily pile up inside the companies already best placed to grab them.

That, Daco argues, is how technological revolutions usually work at first. He points to the railroad boom in the late 1800s and the dot-com era in the 1990s, when large, vertically integrated firms captured most of the upside while smaller rivals were stuck with higher costs, policy uncertainty and pricier borrowing. AI, on that reading, is less a leveler than a sorting machine.

The numbers he cites from 2026 make the case harder to dismiss. Economic output rose 1.7% in the second quarter on just 0.3% more hours worked. Compensation increased 2.6%, but with oil-driven inflation in the spring and summer, Daco said that amounts to flat pay at best, and a slight real decline at worst. Meanwhile, margins reached a record 14.9% of GDP.

The labor share tells the same story from the other side. It fell to 52.8%, the lowest level since the government started keeping track in 1947. Daco warned that 50% is not some natural floor, which is a nice way of saying workers may not have hit bottom yet. If AI keeps amplifying productivity without broadening bargaining power, the “great equalizer” will look a lot like another round of capital winning and labor watching from the sidewalk.