Wall Street tends to reward CEOs who slash jobs in the name of efficiency.
However, those companies are leaning into an area with the “smallest level of return,” Dan Diasio, EY’s global consulting AI leader, told Business Insider.
“There’s a very clear limit to the amount of value you can create by just focusing on productivity and cost reduction,” said Diasio, who also serves as EY’s Americas consulting CTO.
He said AI may be able to shrink the time spent on a task, but it doesn’t necessarily eliminate the job around it. If cost reduction is the main strategy, those companies are effectively asking AI to pay for itself by reducing headcount.
Diasio doesn’t discount that AI can be a powerful tool for cost reduction and increased productivity, especially in areas like software engineering. At EY, Diasio said that AI has enabled engineers to move much faster and, in some cases, has begun merging multiple engineering roles into a single job.
Diasio added that it’s understandable that CFOs want a clear return on technology spend and AI investment. After all, AI doesn’t come cheap: Amazon, Microsoft, Meta, and Google plan to spend as much as $725 billion in capital expenditures in 2026, according to first-quarter earnings calls. That’s roughly $100 billion more in total than originally projected for the year.
Leaders focused only on efficiency may also find that they still need people to run, supervise, improve, and govern AI-enabled workflows. If too much is automated without human context, the output can become generic, brittle, or disconnected from what customers and employees actually need, Diasio said.
However, productivity gains and cost savings are related, not identical. Software engineering clearly demonstrates that distinction, he said.
AI tools speed up coding, drafting, analysis, research, testing, summarization, and workflow routing. However, the role still involves judgment, architecture, coordination, exception handling, client or user context, review, and overall ownership, Diasio said. Companies may see productivity gains before they see structural cost reduction, he said.
Diasio added that leaders cannot ignore the full cost of AI-enabled workflows. These tools still require platforms, governance, human oversight, adoption, maintenance, and recovery when things go wrong, he said. A working AI tool is not automatically an AI tool that pays for itself, the leader said.
He added that when companies approach AI with fear-based strategies, they risk stifling employees’ creativity. In the age of AI, many employers have ramped up the pressure on workers with AI-related layoffs or AI leaderboards tied to evaluations. Diasio said that if workers believe AI is only a head-count weapon, they’re less likely to help redesign the work.
Diasio said AI’s bigger opportunity lies in driving growth by creating new business possibilities. Diasio said that potential “has no limit,” and the harder question to ask isn’t whether AI can perform a task, but whether the organization can redesign workflows around that capability, he said.
“The real promise is about unlocking new things, and that can be new businesses, that can be new markets, that can be new business models for companies,” Diasio said.
When leaders focus on growth and reimagining workflows, AI isn’t something to fear, but an exciting prospect, Diasio said. The leader added that companies have an opportunity to “reimagine an entirely new way of working.” For example, he said if AI gives people time back, the leadership question is where that time gets redeployed.
“Those are the companies that are positioning themselves to win,” Diasio said.
