CEO Survey: AI becomes core infrastructure

The story of AI has been one of momentum, but this year marks a clear turning point. After years of pilot projects being pressure tested amid regulatory uncertainty, AI is now being deployed everywhere. It’s underpinning how organizations build products and how those products deliver value to users. And it’s core to how companies operate and scale.

In our latest survey, a large majority of CEOs reported a sharp rise in AI integration across core product offerings and nearly universal AI use in software development — clear signs that adoption has progressed from experimentation to company-wide transformation. What’s more, CEOs said their investments in AI across various use cases have delivered tangible gains. The next phase will be less about whether to adopt AI and more about how to organize around it. 

Here’s a more detailed look at how CEOs are reshaping operating models, skills, and roles for a world where AI is a structural layer. Explore the full survey results, segmented by region, sector, and stage on Sōzō Pulse.

CEOs report deep product and operational integration

A year ago, just over half (55%) of CEOs said AI was embedded in their core product or service. That figure has climbed to 77%, a dramatic 40% jump that underscores how quickly AI has been embraced. 

“Masa has predicted for years that AI would become the core engine of innovation and human progress,” said Alex Clavel, CEO of SoftBank Vision Funds. “That vision is becoming reality faster than many would have expected.”

AI is now instrumental not only to product development but also to how CEOs build and run their companies: usage jumped to 74%, from 71% a year ago, in operations; to 67%, from 58%, in customer engagement and experience; and to 47%, from 31%, in sales and business development. Just 2% of CEOs said they make no meaningful use of AI in their organizations.

It’s no surprise that when asked what excites them most in the coming year, many CEOs pointed directly to AI. Respondents across industries singled out AI agents, the scaling of internal AI projects, and the expansion of AI use cases as some of the most anticipated developments for the coming year. A fintech CEO said that the rapid shift in AI adoption and its transition out of the experimentation phase are key drivers of his optimism for the coming year.

Why product-led AI strategies deliver clear value

Tom Davenport, an expert in analytics and AI with distinguished academic and industry roles spanning Babson College, MIT, Brown, and Deloitte, sees the increased focus on products and innovation as a promising bet, and one that’s more valuable than focusing on individual productivity. “Individual productivity gains are elusive — they typically come in small quantities [and] are difficult to measure,” said Davenport. “If AI is used to develop new products and services, there are typically measures of how much they are used by internal or external customers, and how much value for customers and providers that they generate.”

For companies deepening their use of AI, Davenport said the shift to enterprise-level initiatives should include “the planned redesign of workflows, with careful attention to the roles of humans and machines in tasks and processes.” Davenport also suggests that leaders should focus on value-led business transformation to get true strategic benefit from AI.

Software development emerges as AI’s most effective use case

Our survey confirms what many industry observers have noted: Software development is quickly emerging as AI’s killer app.

CEOs report that software development is the No. 1 use case for AI, as 84% of companies use the technology for coding. And it’s the AI application with the most consistent ROI, with 78% of companies reporting gains from its use — far outpacing AI adoption in products, which has delivered positive ROI for 57% of companies.

“Coding is the kind of highly structured, rule-bound, and voluminously documented discipline that AI is suited to and that humans struggle with,” said Randall Stross, an emeritus professor of business at San Jose State University and the author of several books on entrepreneurship in Silicon Valley. “It’s no wonder that it’s being adopted so broadly and speeding up software development in the process.”

Early-career engineering roles shift as AI accelerates coding

CEOs report that as the application of AI across use cases is helping drive efficiency, it is also having an impact on their staffing needs: 79% said AI use has already reduced or will reduce their planned headcount. 

The impact of these reductions is likely to be felt most acutely among early-career professionals: 40% of CEOs said their use of AI in software development means they will need fewer junior engineers; 40% also said that AI use in general is decreasing early-career opportunities at their companies. However, 5% of CEOs said AI use has increased their need for engineers.  

The divergent data points help explain why a majority of CEOs believe the jury is still out on AI's ultimate impact on the labor market: 52% said it's too early to say if AI will be a net positive or negative on jobs.