The investigation

Companies, how to bridge the skills gap in the face of artificial intelligence

Wibo research highlights how companies struggle to bridge the gap between current and needed skills

by Gianni Rusconi

24 March 2026

5' min read

Translated by AI

Versione italiana

5' min read

Translated by AI

Versione italiana

The artificial intelligence is profoundly transforming the way companies operate, make decisions and manage work. But there is a paradox that is characterising this evolutionary process: while technology is accelerating, the preparation of organisations is proceeding slowly. This is not due to a lack of tools, but rather to a deficit of knowledge and cultural models suited to accompanying (and managing) the change dictated by the large-scale adoption of new-generation digital solutions.

This is the picture that emerges from the Wibo Future Skills Research 2026, a survey conducted on a qualified sample of managers and HR, Learning & Development and People & Culture managers to assess companies with respect to the 'capabilities' considered strategic in the coming years. The summary figure is emblematic: the average level of skills perceived in organisations today is 5 out of 10, compared with a level considered necessary of 8.9 out of 10. A gap of almost four points that highlights a structural tension between awareness and capacity for action.

"The problem is not how much is invested, but where and how," observe Tommaso Seita and Alessandro Busso, founders of Wibo, a Turin-based innovative start-up active in corporate learning. "Italian companies have never stopped training, but too often they have done it in the wrong direction: catalogue learning, isolated events, classroom days disconnected from operational reality. Thus training becomes a patch, not a strategic lever'. The reflections arising from the research are part of a context of accelerated job transformation, well photographed by the World Economic Forum's 'Future of Jobs Report', according to which in the next five years 39% of the skills currently considered 'core' will become obsolete, while the adoption of AI and automation will generate around 170 million new roles, making more than 90 million of them 'obsolete'. The challenge can therefore not only be technological but also organisational and cultural, and materialises in the need to prepare people to work in an environment where processes, operational tools and responsibilities change rapidly.

The skills that really matter

The nature of the skills prioritised is one of the most interesting passages in Wibo's survey. In the era of artificial intelligence, in fact, the competitive differential does not seem to be represented so much by technical skills as by cognitive and relational ones. When asked which 'attributes' will be most relevant for managers and team leaders in the next few years, the sample indicated complex problem solving in 80% of cases, followed by change management (75%), critical thinking (70%) and the ability to give and receive feedback (70%). Only in fifth place appears a skill directly related to the technological sphere, namely the efficiency of AI-managed processes, cited by 68% of respondents. The resulting message is clear and already widely shared: artificial intelligence and large language model (LMM) technologies do not replace human work, but can amplify its value when appropriate skills come into play (as drivers). 'For years we have considered so-called soft skills as ancillary skills,' Seita and Busso explain, 'whereas in reality they have always been central. Today, AI is dismantling this prejudice, because when a machine is capable of analysing data or generating content, what really remains irreplaceable is the human ability to identify the right problems, make decisions under uncertainty and guide people through change'.

Signs of transformation are also evident on the leadership front. In increasingly hybrid and distributed organisational contexts, skills such as the ability to manage conflict and to process and provide feedback are becoming key elements of team management, and according to 70% of respondents, the latter requirement is one of the most urgent skills to be developed. It is well known, moreover, how a leadership model different from the traditional one is progressively emerging, and Busso has his own precise interpretation of this trend: "The figure of the leader-coach is gaining strength, reflecting an approach of less command and control in favour of more listening, trust and psychological security. Many companies recognise this change on a theoretical level, but still struggle to sustain it in daily practice'.

Generating value is the goal, but technology alone is not enough

The dichotomy between awareness (of the benefits of innovation) and ability to act is evident when it comes to artificial intelligence, perceived as the most disruptive transformative force by some management but also the field in which companies say they feel least prepared. Wibo's research confirms in this regard how the skills required are mainly of an applicative nature: 68% of HR managers, as already reported, indicate as a priority the improvement of process efficiency through AI, while around half emphasise the importance of developing prompt engineering skills, and thus the effective formulation of requests for the production and analysis of content and documents to generative models. "Companies invest in technology, buy licences and launch AI projects, but 95 per cent of the initiatives do not produce concrete results," Seita recalls, pointing out that the limitation is not of the technology as such but due to the fact that "people do not change their behaviour and processes to actually use it".

According to the founders of Wibo, more specifically, artificial intelligence is changing the so-called 'grammar of work'. If in the past the value lay in mastering the process that led from a problem to the solution, today many intermediate steps are being automated and consequently the value is moving upstream, on the ability to identify relevant problems and formulate the right questions. "We talk about problem economy," the two founders point out, "AI makes execution more and more accessible and the competitive advantage is to be found in the ability to identify the right problems, critically evaluate the answers and decide how to use them. In this scenario, technology becomes to all intents and purposes a sort of co-pilot of human intelligence with respect to a model that, in order to function effectively, requires new skills, from problem awareness to the ability to use digital tools appropriately.

From episodic training to 'learning culture'

The issue of skills is inevitably intertwined with that of training, although many organisations continue to perceive corporate learning more as a cost than a strategic investment. And this happens for several reasons. On the one hand, the so-called 'tyranny of urgency', i.e. both the operational pressure that pushes companies to favour short-term actions over medium- and long-term people development plans. On the other hand, the cultural inertia of organisational structures that still remain rigid, where error is penalised and individual initiative is hardly encouraged. 'Both these components,' Seita and Busso observe, 'are part of the same vicious circle. One does not invest in skills because one is overwhelmed by daily operations, but one remains overwhelmed precisely because one lacks the skills to delegate, plan and innovate'.

To get out of this dynamic, this is Wibo's recipe, it is necessary to change perspective, avoiding considering training as an activity separate from work and instead integrating it into organisational processes. It is therefore not just a question of increasing budgets, but of rethinking the way people learn and collaborate through experiential training paths, combining live sessions, coaching and digital tools. The ultimate goal is to make learning immediately applicable in everyday business and the starting point is to observe how people actually work. "It is not enough to map stated competences but we need to start with the daily tasks, understand where time is invested and where it can be freed up thanks to AI. Only in this way,' Seita and Susso recommend, 'is it possible to design training courses that produce real behavioural change. What has worked so far will no longer suffice and those who do not act on the skill gap today will accumulate a knowledge debt that will become increasingly costly to fill'.