This article was originally published on AI News.
The real challenge is not about complying with the new rules, but about AI literacy. The new issue of AI THAT WORKS
Artificial intelligence has moved beyond the experimentation phase and become part of everyday business life. Today it is used to automate administrative tasks, support customer service, analyze data, create content, improve production processes and assist people in their decisions. In many cases it has already become an ordinary component of work, often without organizations being fully aware of how it is used and what implications it carries.
This is the context in which the European AI Act is gradually coming into force. For months the debate focused mainly on regulatory aspects, risk categories and the obligations set out in the regulation. Today, the perspective is changing. The question is no longer what the rules are, but how to make sure they become a tool that supports innovation rather than slowing it down.
For businesses, this means facing a transformation that goes far beyond compliance. Adopting artificial intelligence requires rethinking organizational processes, skills and responsibilities. Governing AI means knowing which systems are in use, understanding how information is processed, defining who is responsible for overseeing the results produced by algorithms, and ensuring that final decisions always remain attributable to human responsibility. This is the real change introduced by the AI Act: a new governance model.
From compliance to governance
Of course, the starting point is not the same for everyone. Large companies, especially those operating in regulated sectors such as finance, healthcare, energy, telecommunications or public administration, already have structures dedicated to compliance, risk management, cybersecurity and data management. For these organizations the AI Act is certainly a significant commitment, even though it fits into processes that, at least in part, already exist.
The situation is different for most small and medium-sized Italian businesses, which form the heart of our production system and often lack the dedicated resources or skills to undertake a journey of this kind. And this is precisely where one of the most important games is played. SMEs are probably the ones that could gain the most significant benefits from artificial intelligence: automating repetitive tasks, improving document management, optimizing customer relationships, analyzing sales data, supporting business decisions or increasing people’s productivity.
However, if regulation is perceived as a set of complex obligations designed exclusively for large organizations, there is a risk that many companies will decide to postpone AI adoption or, conversely, use it without adequate awareness of the associated risks and responsibilities. Both scenarios would hold back innovation, just when the Italian production system needs to accelerate its digital transformation.
Proportionality as the key to adoption
For this reason, the principle of proportionality plays a central role. Not all companies play the same role within the artificial intelligence ecosystem, and not all of them have to meet the same obligations. There is a substantial difference between those who develop AI models, those who build vertical applications, those who integrate third-party solutions and those who, more simply, use widely available tools on the market to increase internal productivity.
Today many companies use platforms such as Microsoft Copilot, ChatGPT Enterprise or Google Gemini to support everyday tasks such as writing, document analysis, information search or content production. In these cases the issue is not building an artificial intelligence model, but using it consciously, by defining internal rules, verification procedures and oversight criteria.
This is exactly where the success of the AI Act will be measured. Companies will need simple guidelines, operational models, self-assessment pathways, practical tools and continuous training. Regulation can truly foster innovation only if it manages to be understandable and applicable in everyday practice, without turning into yet another layer of bureaucratic complexity.
The role of AI vendors is changing
The entry into force of the new regulatory framework will also have a significant impact on the supply side. Software houses, system integrators, consulting firms, startups and providers of AI-based services will no longer be called upon solely to develop innovative solutions. Companies will look for partners able to support them throughout the entire adoption journey, from assessing use cases to managing risks, from documenting systems to training people, through to monitoring the performance and impact of the solutions implemented.
This means that compliance is no longer just a regulatory obligation: it becomes an element of value. Companies will increasingly choose vendors able to demonstrate reliability, transparency and expertise in artificial intelligence governance. The ability to design solutions that comply with European rules will therefore become a competitive factor and a distinctive element of the offering.
The real challenge is called AI literacy
Alongside the technological and regulatory aspects, there is a topic set to become increasingly important: skills. Talking about AI literacy means going far beyond the technical training of specialists. Artificial intelligence skills will need to involve chief executives, boards of directors, HR managers, legal functions, marketing, procurement, IT and all those roles that, in various ways, will make decisions supported by AI systems.
Technology, in fact, does not remove people’s responsibility. On the contrary, it makes the ability to interpret the results produced by algorithms, understand their limits, identify possible biases and assess their impact on business decisions even more important. Without a widespread culture of artificial intelligence, it will be difficult to turn the potential of the technology into a real competitive advantage.
Trust as the engine of innovation
Europe has chosen to build its model for artificial intelligence development by focusing on trust. It is a choice that, if accompanied by balanced implementation and constant dialogue with the business world, can become an important competitive factor. Trust, in fact, is not an obstacle to innovation but one of the conditions that make it possible. Companies, citizens and public administrations will adopt artificial intelligence only if they are convinced that it is used in a transparent, secure and responsible way.
For this reason, the real challenge of the AI Act is not simply to verify compliance with new regulatory obligations. The challenge is to build an ecosystem in which technology, governance and skills grow together, allowing companies to innovate without giving up responsibility. It is on this balance that the competitiveness of Italy and Europe will be decided in the coming years. Because the success of artificial intelligence will depend not only on the quality of algorithms, but on organizations’ ability to integrate them into their processes with method, awareness and a long-term strategic vision.
