As every year, I watched Davos from a distance, but with a lot of attention. The conversations happening in those corridors, on the panels, in the halls, and in the side meetings are shaping what we’ll be doing for the next eighteen months. When your Global Chairman stands up and explains that 56% of companies say they are getting no significant financial benefit from AI so far, you pay attention.
For those who don’t follow it regularly, Davos is the yearly gathering for the World Economic Forum Annual Meeting in Davos, Switzerland. About 2,500 leaders converge in January: CEOs, presidents, investors, technologists, policy makers. On the surface it looks like an expensive conference. In practice, it is where the issues that end up on board agendas are stress‑tested in public: AI, geopolitics, cyber, trade, climate, capital flows. The way those topics are framed in Davos tends to influence how leaders think, what they choose to prioritize, and where they decide to spend their next euro of investment.
Our 29th Global CEO Survey dropped at Davos, based on 4,454 CEOs across 95 countries and territories. The headline numbers are straightforward: CEO confidence about revenue growth in the next 12 months sits at 30%. That’s a five‑year low. Last year it was 38%. In 2022 it was 56%. On AI specifically: only one in eight (12%) CEOs say AI has already delivered both cost savings and revenue gains. Meanwhile, 56% report seeing no significant financial benefit from their AI investments so far.
Those are the facts. But the interpretation matters more than the numbers themselves. What the data is saying is that we have a very large group of organizations that have committed resources, hired people, launched initiatives, and have little or nothing to show for it yet. That creates pressure. On boards, on budgets, on CIOs, on teams.
At Davos, Mohamed Kande addressed this directly. His message was that a clear separation is emerging. Some companies are already turning AI into measurable financial returns. Many others are still in pilot mode, and they’re running out of time to move beyond it. As he put it, “2026 is shaping up as a decisive year for AI. A small group of companies are already turning AI into measurable financial returns, while many others are still struggling to move beyond pilots.” In an interview on the sidelines, he went further: “Only 10% to 12% of companies report seeing benefits on the revenue or cost side, while a staggering 56% say they are getting ‘nothing out of it’.”
He also gave a very clear diagnosis of why: “Somehow AI moves so fast … that people forgot that the adoption of technology, you have to go to the basics,” pointing to clean data, solid business processes, and governance as the missing pieces. He did not describe this as a technology gap. He described it as an execution and leadership gap. The distinction matters because it suggests the companies losing ground aren’t losing because they hired the wrong data scientists or chose the wrong models. They’re losing because they haven’t built the operational infrastructure to turn models into business value.
Why the returns aren’t coming (yet)
I’ve been thinking about this since Davos, and it maps to what I’ve seen in my own work. The companies in that 12% that are getting real returns from AI are not categorically smarter. But they are doing something systematically different.
They started with the business outcome, not the technology. Before building anything, they asked: what financial KPI are we trying to move? What measurable difference do we want in six months? That sounds obvious, but most organizations don’t do it. They build first and figure out the business case after.
They invested in the platform layer. Not just one model, but the infrastructure that lets them deploy models, monitor them, iterate on them, and scale them. MLOps. LLMOps. The infrastructure that most tech teams see as unglamorous work and try to avoid. Except you can’t scale AI without it.
They built governance from the beginning. Responsible AI frameworks. Clear ownership of outcomes. Risk management. Explainability. Again, this sounds like work that should come later. But the companies getting returns built it in from the start. And they connected IT to the business in a real way. Not IT implementing what business asks for, but business and IT jointly owning the outcome. That alignment changes the conversation because it forces clarity on what success actually means.
The survey actually quantified this: organizations with strong AI foundations—Responsible AI frameworks and technology environments that enable enterprise‑wide integration—are three times more likely to report meaningful financial returns than those without them. That’s not marginal. That’s structural.
The confidence drop would be worrying enough on its own. But there’s a second dimension to what’s happening, and it’s making the execution challenge harder. Cyber is now a board‑level issue. 31% of CEOs cited it as a major threat in 2026, up from 24% last year and 21% two years ago. 84% say they plan to strengthen enterprise‑wide cybersecurity as part of their response to geopolitical risk. That 84% number tells you something has shifted. Cyber moved from IT concern to existential concern in the minds of boards.
Which means for CIOs, we’re facing an equation that’s gotten more complicated. We need to accelerate AI deployment and harden defenses. At the same time. With the same resources. The tension in that equation is real, and I heard it reflected throughout the WEF coverage.
So what am I taking from all of this? A few things that I think are worth being clear about.
The pilot phase is functionally over. Organizations that are still primarily testing and learning are now in competitive trouble relative to those that have moved to scaling. If we’re not asking ourselves how to take what works and deploy it at scale, we are behind.
The returns won’t come from better models. They’ll come from better operational foundations. That means committing to governance frameworks that actually work, not just check boxes. It means fixing data quality, which is boring and expensive and necessary. It means building infrastructure that lets you iterate, not just pilot. It means clear ownership of financial outcomes, all the way up.
We need a strategy that accelerates AI and hardens defense simultaneously. That’s the difficult part. That’s also the differentiator. The competitive window is real. The companies that have already industrialized AI are pulling further ahead. The penalty for staying in pilot mode is compounding with every quarter.
The real opportunity
What struck me most about the WEF coverage wasn’t the pessimism about revenue confidence or the scale of AI returns that haven’t materialized. It was that this situation creates an opportunity for CIOs. Not because we know neural networks better than anyone else. But because we understand how to build operational foundations. We understand governance. We understand how to connect technology to business outcomes. We understand how to build at scale.
That’s what’s actually missing in most organizations right now. That’s where the gap is between the 12% getting returns and the 56% getting nothing. So we have a moment. The board is paying attention. The strategic importance of getting AI right is no longer abstract. The companies that move decisively on this -that build the foundations, connect the outcomes, move with discipline and confidence -will look very different from the ones that stay in pilot mode.
That’s what I’m thinking about as I head into the next phase of the year. How we make sure we’re in the first group, not the second.
