The hidden AI risk for workers isn't just unemployment — it's a pay cut, former Salesforce AI CEO says

• Former Salesforce AI CEO Clara Shih says AI may hurt workers through lower pay, not just layoffs.
• Wage cuts are a more common way new tech hits workers, she said.
• This happens as technology lowers skill requirements and displaced workers move into new fields, she added.

Much of the debate about AI's impact on the labor market focuses on whether — and how many — jobs it will eliminate.

But Clara Shih, the former CEO of Salesforce AI, says the bigger risk for many workers is actually lower pay.

"While full AI role displacement will happen in certain roles, history shows that wage resets are a more common, insidious, and often equally disruptive way that new technologies affect workers," Shih wrote in an X post on Sunday.

Three ways AI could push wages down

Shih outlined three ways new technologies can lower wages.

One is what she called an "intra-sector squeeze," where workers who lose jobs in an industry compete for the remaining roles in the same field, pushing wages down.

Shih pointed to manufacturing after the early-2000s trade shocks as an example. As factories closed or automated and manufacturing moved overseas, laid-off workers fought for a shrinking pool of domestic manufacturing jobs, and real wages declined, she said.

The US Bureau of Labor Statistics estimates that 5.5 million US manufacturing jobs were lost between 2000 and 2017.

In a 2016 paper for the National Bureau of Economic Research, economists David Autor, David Dorn, and Gordon Hanson found that workers in industries exposed to import competition from China "accumulate substantially lower earnings" between 1992 and 2007.

Another dynamic Shih cited is that technology can lower the skill barrier for previously specialized work, expanding the labor pool.

"AI (like past tech waves) slashes the skill floor for once-premium jobs, flooding labor supply and compressing wages," Shih wrote.

She cited London's black cab drivers as an example. For decades, drivers had to master "The Knowledge," a rigorous examination process that required memorizing thousands of streets and landmarks.

But GPS navigation and ride-hailing apps dramatically reduced the need for that expertise and expanded the driver labor pool, exposing drivers to more competition.

A third factor involves workers moving into entirely new sectors after losing higher-skill jobs.

"Displaced high-skill workers switch fields, often taking a pay cut while displacing incumbent workers," Shih wrote.

For Shih, the implication is that policymakers and workers shouldn't judge AI's labor-market impact solely by job losses, but also by wage trends.

The early AI wage boost may already be starting to fade

Ioana Marinescu, an associate professor at the University of Pennsylvania School of Social Policy & Practice and coauthor of a recent Brookings Institution paper on what she calls "intelligence saturation," told Business Insider that AI may already be nearing the peak of its wage boost.

New technologies often raise pay early on by making workers more productive, she said, but that effect can reverse once automation spreads widely enough.

Her model suggests wage growth could begin to decline once roughly 37% of cognitive or "intelligence" tasks are automated — a tipping point where automation starts replacing, rather than augmenting, workers.

By her estimate, the economy has already automated more than 14% of those tasks — meaning the peak of the AI-driven pay boost may arrive sooner than many expect.