The 93% AI Spending Gap Quietly Tanking Worker Output

Corporate America is pouring money into artificial intelligence infrastructure and almost none into the workers expected to use it. About 93% of enterprise AI investment targets the technology itself, while just 7% goes toward preparing the workforce for the shift, according to data cited by Simona Spelman, U.S. Human Capital Practice leader at advisory firm Deloitte.

That lopsided spending ratio carries a direct financial penalty. Deloitte’s research shows that when employees distrust their leadership across four specific metrics—reliability, transparency, capability, and humanity—only 7% remain motivated to perform at their best. Even a single failure in one category slashes motivation from 82% to 37%. Full trust across all four factors pushes that figure to 96%, Spelman said during Fortune’s Most Powerful Women Summit in Santa Barbara, Calif., last week.

The disconnect between executive planning and worker reality suggests companies are treating AI deployment as an IT procurement exercise rather than an operational overhaul. Jennifer McCollum, president and CEO of workplace inclusion nonprofit Catalyst, said the group’s research found 68% of leaders claim they have reskilling plans in place. Only 34% of employees agree those training programs are actually accessible to everyone.

“Our intentions are good, but we can really get it wrong,” McCollum said.

Catalyst found only one-third of workers believe their leadership teams act with integrity during business disruption. A mere 16% said executives remained confident and supportive specifically during AI-driven change.

This credibility deficit hits unevenly. Forthcoming Catalyst research indicates women face a distinct double bind: using AI tools at work invites questions about their credibility, while refusing to use them triggers doubts about their capabilities, McCollum warned.

The way executives communicate new software rollouts frequently undermines their own productivity goals. Francesca Cornelli, dean of Northwestern University’s Kellogg School of Management, noted that framing AI as a tool to simplify processes often backfires.

“All they hear is, ‘you won’t need me anymore,’” Cornelli said.

Cornelli argued that as knowledge becomes democratized through these systems, the technology alone will stop functioning as a competitive differentiator. What remains is “investment in the people, in the human capital.” But verbal commitments without corresponding budget shifts fail immediately. “If you just use words, people will see through it,” she said.

Spelman pointed out that once the initial software advantage fades, businesses will be separated by “enduring human capabilities like empathy, curiosity, and judgment.” Those traits require funding to develop. They cannot be installed via a software update.