In mid-2026, PwC went back to a group of chief executives it had polled eight months earlier and asked a short set of questions. Confidence in revenue growth held about where it started. That was the headline PwC ran with. The number worth your attention sits a little lower in the data.
About 39% of CEOs told PwC their companies maintained or improved a positive AI impact. Many leaders are still working out how to turn AI into reliable results. So the picture at the top of the house is not one where every company is already winning with AI. Many leaders are mid-climb. If you count yourself among them, you are in ordinary company.
For a leader who owns a P&L, the useful question is what separates the companies where AI pays off from the ones where it does not. PwC's data points at an answer. It is not the size of the AI budget.
What did PwC's mid-year CEO survey find?
PwC surveyed 351 CEOs across 59 countries and 27 industries between May 15 and June 22, 2026. It was a short follow-up with leaders drawn from its 29th Global CEO Survey. Growth confidence held near where it started. AI results split the group into performers and strugglers.
The full report is titled "CEO confidence holds steady despite geopolitical disruption and rising costs." That title carries the whole plot. Costs went up. Decisions got harder. Confidence did not crack. Most CEOs reported that energy and other input costs rose because of global shocks in 2026, and the group stayed steady anyway.
One thing to keep straight before you quote any of this in a board deck. These are poll findings, not audited results. They record what CEOs report about their own companies. A PwC survey of executive opinion is a good read on sentiment. It is not the same as a measured return.
Why did CEO confidence hold steady in 2026?
Roughly 42% of CEOs said they were very or extremely confident about revenue growth over the next 12 months, close to the 39% who said so eight months earlier. More reported rising confidence than falling: 33% up against 26% down. That balance held even as most saw costs climb.
Context makes the steadiness read better. PwC's 29th annual Global CEO Survey, released in January 2026, was built on 4,454 CEOs across 95 countries and territories. In that earlier read, confidence in the revenue outlook had dropped to a five-year low. So holding steady in mid-2026 describes a group that stopped sliding, well short of a boom.
And they stopped sliding through a rough stretch. Geopolitical disruption. Rising energy and input costs. Strategic calls that got more complicated, not less. Steady, under those conditions, is its own kind of signal.
What is techno-resilience, and why does it predict AI success?
PwC introduced a construct it calls "techno-resilience." It combines three things: long-term thinking, resilience capabilities, and strong AI foundations. Companies high on techno-resilience were 74% more likely to report AI success than companies low on it, 55% versus 32%. They were also 66% more likely to hold strong confidence in future revenue, 48% versus 29%.
Companies high on techno-resilience were 74% more likely to report AI success than companies low on it, 55% versus 32%.
Source: PwC CEO Survey Mid-Year Snapshot, 2026Sit with the size of that difference. When PwC lined up the companies strong on techno-resilience against the ones weak on it, better than half of the strong group reported AI success while roughly a third of the weak group did, and the confidence spread ran nearly twenty points wide across the two groups. That is a large delta for a single survey.
A leader can act on this part. The difference PwC found does not track who bought the most AI. It tracks who built the capability around it. Long-term thinking is a leadership habit. Resilience is an organizational one. AI foundations are a team skill. None of those three is a purchase order. All three are things you own as an executive.
That is the honest read for anyone who feels behind. The returns are not going to the biggest spenders. They are going to the companies that built the muscle to use what they bought.
The returns are not going to the biggest spenders. They are going to the companies that built the muscle to use what they bought.
How does team capability turn into AI results?
This is where the 7 Levels of AI Proficiency helps a leader think clearly, because it puts a name on the capability PwC is measuring from the outside.
In the 7 Levels of AI Proficiency, the entry point is Level 1: The Cadet (AI Aware). That is a person who has tried AI and knows it exists. Useful, but early. The technical rungs climb from there through prompting and critical thinking and context management. What surprises most executives is where the framework tops out.
The top three levels are about people, not prompts. Level 5: The Captain (Design Thinker) designs AI for others to use, not just for personal output. Level 6: The Admiral (Systems Integrator) turns one-off wins into documented processes that produce the same quality every time. Level 7: Mission Director (AI Orchestrator) leads the organizational change: culture, trust, and the safety people need to work differently. The reason those sit at the top is human skill, not technical skill.
Read PwC's techno-resilience against that. A company strong on techno-resilience looks like a company whose leaders have moved past personal experimentation and into designing how AI runs across teams. That is what the 55%-versus-32% difference in AI success is describing from a survey seat. Capability, built deliberately, showing up as results.
For a P&L owner, this is the useful translation. You are not being asked to code. You are being asked to build a team that can climb. The 7 Levels of AI Proficiency gives you a way to say where your people sit today and where the next rung is.
What does this mean for where you place your AI bets?
Start with the near-term payoff PwC actually measured. Nearly four in ten CEOs, 38%, said their company used AI since January 2026 to spot and act on new business opportunities created by changed conditions, things like changes in customer demand. The use was practical: seeing change earlier and responding faster than the company down the street.
That is a bet you can place without a moonshot budget. The question for your next planning session is narrow. Where does seeing change a week earlier turn into revenue or saved cost? Point AI at that, and you are chasing the same 38% who already found it.
A few things to tell the board, grounded in what PwC found:
- The results are uneven at the top. Only 39% of CEOs report maintaining or improving a positive AI impact. Feeling mid-climb is common, not a red flag.
- Spend is not the predictor. Techno-resilience is. The 74% edge in reported AI success went to companies that paired AI foundations with resilience and long-term thinking.
- Capability is the asset. You build it in your people, and it compounds. A tool depreciates. A team that can design and run AI does not.
One caution, stated plainly. A 55% success rate reflects the strong group's self-report in a sentiment poll, short of a promise. The lesson is directional: build the surrounding capability, and your odds of a real return improve. That is a bet worth making. It is not a guarantee to underwrite.
A next step
You do not need a new AI budget to act on this. You need a clear read on where your team can climb.
Pick one process where seeing change a week earlier would help your P&L. Ask who on your team could design an AI approach for it, and who could run it consistently once it works. That single conversation tells you where your people sit in the 7 Levels of AI Proficiency, and where the next rung is.
The companies PwC scored high on techno-resilience got there by building that capability on purpose. So can yours. Where would one week of earlier warning change a number you report to the board?
Related reading: Level 7: The Mission Director (AI Orchestrator).
Sources
- CEO confidence holds steady despite geopolitical disruption and rising costs (PwC CEO Survey Mid-Year Snapshot)
- CEO Survey Snapshot, August 2026 (PwC)
- PwC 2026 Global CEO Survey: Leading through uncertainty in the age of AI
Frequently Asked Questions
How many CEOs did PwC survey for the mid-year snapshot?
351 CEOs, across 59 countries and 27 industries, fielded between May 15 and June 22, 2026. They were drawn from the larger group that took PwC's 29th Global CEO Survey in late 2025.
Are these AI numbers measured returns or opinion?
Opinion. This is a survey of what CEOs report about their own companies. Treat the 39% positive-impact figure and the techno-resilience differences as sentiment from PwC's poll, not audited financial results.
What is the 29th Global CEO Survey?
PwC's annual study, titled "Leading through uncertainty in the age of AI," released January 2026 and based on 4,454 CEOs across 95 countries and territories. In it, revenue-outlook confidence dropped to a five-year low and AI showed up as a divide between leaders and laggards.
What is techno-resilience in one line?
PwC's mix of long-term thinking, resilience capabilities, and strong AI foundations. Companies high on it were 74% more likely to report AI success, 55% versus 32%.
Find your AI Proficiency level
The free 7 Levels assessment places you across seven stages of AI capability. Under ten minutes. Research-backed scoring.