Italians say they understand artificial intelligence less than almost anyone else in the West. Meanwhile, AI’s productivity dividend is concentrating among the companies that integrate it. For an SME, that gap is already a concrete competitive factor today.
Three numbers, taken from three studies published in 2026, tell a story that should concern anyone running a business in Italy.
The first comes from the Ipsos AI Monitor 2026, the survey conducted across 32 countries. Asked whether they feel they have a good understanding of what artificial intelligence is, only 54% of Italians say yes. It is the second-lowest figure in the world (only Japan ranks below Italy) against a global average of 72%. It is a clear signal: in the country, AI still exists mainly as a media phenomenon rather than as an everyday work tool.
The second comes from PwC’s Global AI Jobs Barometer 2026, which analyzed more than a billion job postings across six continents. Companies in the sectors most exposed to AI record productivity growth 40% higher than the least exposed ones. And since 2022, the year AI use exploded, that advantage has tripled.
The third is the synthesis of the first two. While the productivity of integrated companies accelerates, in Italy the majority of people still do not feel able to explain what this technology actually does. The gap between what AI is producing and what the Italian market perceives is, itself, the risk.
The real game is competitive divergence
Public debate on AI almost always revolves around one question: will AI replace me? In Italy, 28% of workers think it is likely that AI will take their place within the next five years — a figure that is, incidentally, below the global average of 35%. PwC’s data, however, indicate that the decisive dynamic is a different one.
The companies most exposed to AI are growing: they hire more (headcount up 52% versus 36% for the least exposed) and they pay more (wage growth of 24% versus 17%). The organizations that achieve the greatest productivity gains use AI to amplify people. Tellingly, the roles most exposed to artificial intelligence are adding — two and a half times faster than others — tasks that require empathy, judgment and creativity: the human skills that AI makes more valuable.
The real competitive mechanism, then, pits companies against one another. PwC describes a two-speed labor market, and the same logic applies among companies: those that integrate AI into their processes accumulate an advantage, year after year, over those that keep it at the margins.
For an Italian SME, the concrete risk looks like this: a smaller, younger, more integrated competitor gaining efficiency and margins while you postpone the decision.
Silent adoption
There is a second figure from Ipsos that deserves attention, because it describes an Italy already in motion. 65% of Italians say they use AI tools even when they do not fully trust them. And 56% say AI has saved them time at work over the past year.
Translated: artificial intelligence is already inside Italian companies. It came in from the bottom, through individual employees, without governance, without measurement, without a strategy. It is real adoption, still invisible, and for that reason not capitalized on.
This is the costliest blind spot. A company can have dozens of people using AI every day to produce faster, and have no way of knowing whether that gain translates into value, into quality, into margin. The time savings are already there; the next step is turning them into a business result.
Buying tools is the wrong strategy
The instinctive reaction to this scenario is to buy. A new tool, a license, a platform. It is exactly the mistake documented by the data on martech.
According to Gartner, two-thirds of marketers now use sixteen or more technology solutions, yet the actual utilization rate has fallen to 49% of available features, down from 58% in 2020. Only one capability in three, in the average stack, is really used. Abundance without architecture produces friction: every new tool adds a node to a network that already struggles to communicate.
The answer, as McKinsey argues, lies in building a layer of connective intelligence: a layer able to orchestrate the tools that already exist toward a goal. The point is crucial: today the unit of competitive advantage is the system that holds tools and campaigns together and steers them toward a common result.
For an SME, which does not have the budget to accumulate mistakes, the lesson is even sharper. AI tools are now cheap and everywhere. The real lever is integrating them into a process capable of producing a measurable result.
From tool to result
There is a structured way to think about this integration. A four-step ladder sums it up well: tool, co-pilot, autopilot, outcome.
A tool does one thing when you ask it to. A co-pilot works alongside you and makes suggestions. A system on autopilot runs entire processes under supervision. And at the final step, the outcome, it delivers the business result: more qualified visibility, more conversions, more margin.
Most Italian companies stop at the first step, buy tools and wait for the result to arrive on its own. The result, instead, comes from integrating the technology around a clear metric.
This is where the principle becomes operational. The measure generates the work. You start from what you want to achieve and how you measure it; then you integrate AI at the points in the process where it really moves the needle; finally you automate what is repetitive, freeing people for judgment, relationships and strategy: exactly the skills that, according to PwC, AI makes increasingly valuable.
The decision moves upstream
So far we have looked at AI from inside the company, as a capability that amplifies those who integrate it into their processes. There is a second front, mirroring the first: artificial intelligence is also the place where customers form an opinion about a company, often before they have even contacted it.
Anyone looking for a supplier today increasingly opens a conversation with an assistant, not ten browser tabs. They ask the question, receive a synthetic answer, form an impression. The research phase — the one that once required visiting and comparing several sites — now happens inside the dialogue with the model.
From this comes a widespread fear: if the person finds what they are looking for in the AI’s answer, they never reach the website. For a publisher this is a substantive problem (the click on the page is the product, what generates revenue). For a company that sells products or services the calculation changes: the click has always been a step toward the only thing that really matters — being chosen.
Read this way, fewer clicks describe above all a supply chain that is shortening. The information phase moves upstream, into the conversation; the customer comes out of it with a shortlist of names already in mind and, when ready, searches directly for where to buy the brand that solved their doubt. Whoever appears in that answer reaches the first contact with a more informed interlocutor, closer to the decision.
The risk shifts accordingly. The relevant question becomes whether your name is in the answer at the moment the decision is forming. There, presence works almost like a threshold: you are in the shortlist the model builds and you stay in the game, or you disappear from consideration precisely when it matters most. And the shortlist is short, chosen anew each time.
It is the same logic of the result, applied to demand. The outcome to govern is not traffic, but qualified presence at the instant a potential customer is deciding who to turn to: in which answers the brand appears, with which sources, next to which competitors. It is a presence that can be measured, and what can be measured can be governed.
It holds for internal productivity as for external visibility: in both cases the advantage comes from the system that steers the tools toward a result, not from the tool itself.
Two futures, one choice
From this reading of the data comes the way we work at Sequel. We support companies that have the ability to move quickly to integrate artificial intelligence where it shifts a business indicator: in internal processes and in the presence that decides whether a customer chooses them, always starting from a clear result and from how to measure it. We place the solutions at the points where they genuinely move a business indicator, connect them to the tools already in use, and steer them toward a concrete outcome.
The three 2026 studies outline two possible trajectories for Italian SMEs. In the first, AI remains a topic of conversation and a half-used tool, with time savings that struggle to translate into margin. In the second, it becomes an integrated system that produces measurable results and builds a solid competitive advantage over time.
What makes the difference between the two trajectories is integration. And it is a choice made now.
Frequently asked questions
