Cegos International Barometer 2026

Artificial intelligence: work is changing faster than organisations

AI does not eliminate jobs but transforms roles and skills, forcing companies to undergo an unprecedented organisational and cultural transformation

by Gianni Rusconi

27 July 2026

Aggiungi Il Sole 24 Ore
ai preferiti su Google

6' min read

Translated by AI

Versione italiana

6' min read

Translated by AI

Versione italiana

The debate on artificial intelligence in the workplace continues to be dominated, amongst many others, by one question in particular: how many jobs will be replaced by chatbots and agents? Data from the Cegos International Barometer 2026 “Transformation, Skill & Learning’ suggest a broader interpretation of the issue, highlighting another trend: AI is not eliminating jobs but is profoundly changing (and will change even more in the future) their content, the skills required and working methods, forcing organisations to adapt at an unprecedented pace. The survey, carried out among an international sample of HR directors and employees, paints a picture that is, on the whole, already familiar, but with some indicators that are surprising in certain respects. Firstly, 68 per cent of HR managers identify artificial intelligence and automation as the main factor set to affect professional skill sets over the next two years; only 31 per cent of employees fear losing their jobs, whilst 74 per cent are convinced that it is the nature of their role itself that will change. At the same time, HR managers estimate that almost a quarter of existing roles could face skills obsolescence within three years.

Avoiding the risk of a technology-centred approach

The transformation currently under way therefore appears to be gradual and not as disruptive as it is sometimes portrayed, although it is taking place at a deeper level than the public debate suggests. The real issue to reflect on (and act upon), in other words, is not the replacement of people but the ongoing redefinition of their contribution to the business. And there is one finding from the Cegos study that reinforces this point: 74 per cent of workers expect a substantial change in their role by 2035, effectively confirming that the perception of change is now structural rather than episodic. And it is precisely this, according to Emanuele Castellani, Executive Board Member of the Cegos Group and CEO of Cegos Italia, that is the strategic challenge management must address. ‘AI,’ he explained to *Il Sole 24 Ore*, ‘can be an extraordinary catalyst, but its technological foundations only generate value if they are channelled through people. The risk today is to think that transformation hinges entirely on the adoption of innovative tools, on the ability to automate processes or, even more simplistically, on learning to write better prompts. It is a partial view.” Clear words and equally explicit concepts, which take concrete form in a challenge that, according to Castellani, calls on organisations to guide people along a path of conscious engagement, working on mindset, critical thinking, collaboration, responsibility and the management of uncertainty. This complex picture also includes an element that is often underestimated but closely linked to the risk of a technology-centred drift. ‘AI,’ the manager pointed out, ‘is not merely a productivity booster; it can become a new source of constant pressure on people, with implications for stress, relationships and the organisational climate. Senior management must steer artificial intelligence as an organisational and cultural transformation, not merely a technological one.’

People run, organisations chase

One of the most interesting findings from Cegos’s 2026 Barometer is the gap between individual confidence and confidence in the organisation as a whole. Employees rate their own ability to adapt to the world of work in 2035 at over 7 out of 10, whilst they are more cautious in their assessment of their organisations’ ability to keep pace. Other indicators also confirm this perception. For example, 91 per cent of HR managers consider skills development a strategic priority, whilst 77 per cent believe their company is sufficiently agile in responding to training needs. From the employees’ perspective, however, over 40 per cent of those surveyed (in Italia, this figure rises to 53 per cent) state that training is provided too late.

These figures clearly highlight the often paradoxical nature of the so-called ‘time to competency’ and reflect the trend whereby companies have now realised that their competitive advantage depends on the speed with which they develop new skills, yet still struggle to translate this awareness into equally rapid processes. “People – as Castellani sums it up – feel they can change but do not always perceive the organisation as being equally ready. Change does not become systemic if it remains confined to the enthusiasm and initiative of individuals or to isolated experiments, but nor does it work if it is imposed from above in a top-down manner’. Those at the helm of companies, explains the CEO of Cegos in detail, must be able to combine momentum from the bottom up with coordination from the top down, fostering curiosity and experimentation whilst at the same time providing direction, priorities, a safety net and alignment with corporate objectives. The role of leaders therefore becomes crucial, Castellani adds, “because today they must transform individual energy into collective capability, making learning not a one-off event, but a permanent way of working.”

Gen AI is growing faster than its governance

The adoption of generative artificial intelligence – a true watershed moment (with the global launch of the first version of ChatGPT in late 2022) – is progressing rapidly, but continues to develop unevenly. Or at least that is what workers say: 79 per cent of those surveyed confirm that they use Gen AI tools in their personal lives, whilst professional use stands at 64 per cent (and 56 per cent in Italia). Even more significant is the data concerning the gap between actual usage and how the tool is managed: only 32 per cent of employees have received training either within their organisation or through AI-dedicated communities, whilst just 28 per cent of companies (22 per cent in Italia) state that they have formalised and shared rules for using the technology.

What these figures reveal is fairly obvious. In many cases, AI is being integrated into processes even before there are agreed rules on security, ethics and how these tools should be used, giving rise to a twofold risk: on the one hand, the spread of practices such as “shadow AI” practised by employees entirely on their own initiative and outside any corporate governance framework; on the other, a growing disparity in skills fuelled by individual and uncoordinated learning paths. ‘Some employees,’ notes Castellani, ‘are curious, engaged and already using the technology, but this also entails various risks due to a combination of factors, ranging from opportunistic learning to the illusion of mastery of the tool, right through to the absence of an ethical and security framework. The message Cegos is sending to managers is therefore crystal clear: without a framework and clear rules, widespread skills and safe spaces for experimentation, AI risks remaining a collection of individual initiatives rather than a genuine driver of effectiveness.”

Being ‘AI-ready’? It means learning continuously

Finally, one of the chapters of the Barometer focuses on how skills are developed and acquired, noting that 65 per cent of HR professionals prioritise upskilling existing staff over recruiting new candidates, and that 57 per cent focus on internal mobility. This dual indicator appears to be consistent with the 23 per cent of workers who already feel, or anticipate, that their skills are becoming obsolete, confirming that the professional development cycle is shortening significantly. At the same time, the integration of training into day-to-day work is also growing, with 64% of HR managers identifying ‘on-the-job learning’ as a priority approach, whilst nearly two in three organisations are already using (or are about to introduce) generative AI tools into their training programmes. From the employees’ perspective, as many as 67 per cent say they are seeking more meaning in their work than they did three years ago, once again underpinning the principle that technological transformation does not diminish a person’s sense of purpose but can, in fact, enhance it.

The projection for 2035 outlined in the research confirms, in this regard, that the underlying objective for companies should not merely be to teach their staff how to use these new tools, but should above all be aimed at reducing the time between learning and practical application. ‘In an increasingly digital and fast-paced environment,’ concludes Castellani, ‘training must balance performance, employability and the meaning of work. Being AI-ready does not just mean adopting tools, but making the organisation more open, more adaptable and better able to engage in exchange.’ In short, the quality of management (and consequently of the organisation) becomes the true centre of gravity of the transformation driven by artificial intelligence, and companies’ ability to build systems of continuous, integrated and rapid learning is the real competitive factor. The key lies in the awareness needed to reduce the misalignments that technology risks accelerating (between required skills and available skills, or between individual and organisational use of tools) and in the speed with which companies manage to transform AI into widespread expertise.