When it comes to companies and AI, the themes have been consistent. Investments and adoption are rising, but value is not materializing in the P&L. About 88% of companies have adopted AI somewhere in the business, according to Stanford’s 2026 AI Index, showing that adoption is strong. Yet, a recent PwC CEO study found that 56% of CEOs had seen neither revenue growth nor cost reduction from AI, and only 12% have seen both. At the same time, IDC estimates enterprises will spend $409 billion on AI platforms, applications, and services this year. That is a lot of investment for a technology that most companies are struggling to translate into meaningful value.
Why, after several years of investment and attention, are so many companies still not getting value from AI?
CEOS DO NOT UNDERSTAND AI
So far this year, I’ve spoken with roughly 50 CEOs, board members, and business unit leaders from across industries and geographies, at companies ranging from the mid-market to the Fortune 50. From these discussions, two points stand out.
First, CEOs know AI matters. Clearly, awareness is not the problem.
Second, many still do not understand it well enough to form their own view of what it could mean for the business. So, they delegate responsibility for the company’s AI agenda to the CIO, a new AI leader, a business unit leader, or an operating team.
That is reasonable. With a broad mandate, much of a CEO’s job involves deciding what to delegate and to whom.
AI VALUE IS NOT TIED TO OUTCOMES
The challenge comes when CEOs delegate before they understand enough to set direction and align the leadership team around where AI should create value. Without that foundation, they are also less able to judge the plans, roadmaps, and investment proposals their teams develop. With AI, I see that happening far too often. This situation manifests across several common patterns: AI-led pilots without clear business ownership, roadmaps with activities but no or only loosely defined outcomes, and initiatives not aligned to the P&L or supported by corresponding budget commitments.
This scenario does not play out in all organizations. However, elements of it are common when leadership teams, starting with the CEO, cannot confidently articulate how AI can improve the economics of the business.
The numbers back it up. Many AI initiatives stay disconnected from how the business runs and what it’s worth. A recent BCG survey found that just 26% of CEOs have included AI in wide-ranging business transformation. Only 14% of CEOs clearly define their AI initiatives’ P&L impact.
In private equity, the lack of alignment cost can be even higher, because time is part of the value equation. If a portfolio company spends a year deciding who should own AI, waiting for an internal roadmap, or building expertise from scratch, a meaningful portion of the hold period is gone before the work begins. In a five-year hold, one year is 20% of the clock.
What is striking is how differently many companies treat AI from other major investments. Plant and equipment, operating improvements, acquisitions, and other strategic investments would rarely be approved with such a limited connection to expected financial outcomes.
WHAT CEOS NEED TO UNDERSTAND FIRST
Realizing value from AI requires a personal investment in learning, not technical fluency. CEOs need to understand how and where AI can move the P&L, which economic or operational lever AI can improve, and the decisions and workflows where these levers can be best applied.
Consider a logistics company driving 50 million miles annually at the current industry average operating cost of roughly $2.34 per mile, according to ATRI. That represents roughly $117 million in annual operating expense. One lever is miles traveled; the workflow is routing. If better AI-based routing reduced miles traveled even by 5% to 10%, that would create $6 million to $12 million in gross operating-cost opportunity. How much reaches the P&L depends on which costs can be removed or redeployed.
The CEO does not need to design the routing model, but does need to understand enough about AI and the business to see the connection between a capability, a workflow, the economic lever it moves, and the business value it creates. With that understanding, delegation is appropriate.
Realizing value from AI requires a personal investment in learning. CEOs who are getting serious about AI are spending more time with the technology, talking with peers, investing in executive education, and bringing in outside experts to help get them up to speed. The objective is to know enough to ask the right questions, challenge the answers, and recognize whether the company is headed in the right direction. That is the gap worth closing.
Todd James is founder and CEO at Aurora Insights LLC.