Cybersecurity, trust drives the era of artificial intelligence

Trust, even before technology, is the starting point when cybersecurity enters the most exposed and least structured organizations: small and medium-sized enterprises, local administrations, and healthcare organizations. Artificial intelligence is no longer merely generative; it is moving toward agentic forms capable of influencing both defense strategies and the ways attacks are carried out. In this scenario, the response cannot be reduced to the purchase of IT tools, because tools, even when advanced, require interpretation, prioritization, and operational responsibility. AI applied to cybersecurity can already produce highly significant effects: some models designed specifically for this field can shorten the detection of zero-day threats from months to minutes. This kind of capability changes the scale of the problem and makes artificial intelligence an extremely powerful tool. The same power, however, can also be used by attackers. This is why AI should not be regarded as an automatic and definitive answer. Artificial intelligence may appear assertive, but that does not mean it is always accurate, specific, or reliable. Human expertise therefore remains essential, not as an accessory, but as a condition for turning the output of an automated system into a useful decision. If an AI-based platform reports one hundred or one thousand vulnerabilities, but the company does not have the time, people, and resources to understand and fix them, the volume of information risks failing to produce any real improvement. The value lies in the ability to read those results, distinguish what is urgent from what can be planned, and intervene where the risk is most concrete. For this reason, ignoring artificial intelligence would be a mistake, but entrusting the entire process to it without skills and experience would be just as fragile. Defense arises from the integration of technology, method, and trained people: AI must be used, but made precise and reliable through the work of professionals capable of governing it.

Sustainable measures for organizations with limited resources

Many organizations rarely have large budgets or internal departments dedicated exclusively to cybersecurity. Precisely for this reason, the issue is not to import models designed for large infrastructures and apply them without adaptation, but to identify proportionate interventions. For an SME, a local authority, or a healthcare facility, the priority is to lower risk to an acceptable level with measures that are sustainable in terms of cost and time. The approach to adopt is close to the medical one: there is no need to administer an indefinite number of treatments; what is needed are the right ones for the situation. Applied to cybersecurity, this means identifying the minimum measures that make it possible to protect the organization from a significant portion of attacks, without imposing disproportionate investments or processes that are impossible to maintain. Cybercrime is constantly looking for new targets; a company, therefore, cannot afford to remain completely exposed. The starting point is user awareness. The human factor weighs heavily: training cannot be exhausted in half a day, because the attention required of employees and collaborators must be nurtured over time. The goal is not to turn every worker into an IT specialist, but to maintain a strong ability to recognize suspicious signals and risky behavior. The decisive moment is when the click to be avoided arrives: if training works, that is where it produces its effect. To achieve this, repeated campaigns, simulations, and activities built around the organization’s actual habits are needed, not abstract exercises disconnected from daily work. Alongside awareness, there is endpoint protection, meaning the computers and devices used every day. The market offers many valid products, but the choice must be informed and proportionate, avoiding improvised or unreliable solutions. Another key element is the verification of defenses through vulnerability assessments and penetration tests. Even highly skilled developers and system administrators can leave a poorly configured port, an outdated library, a vulnerability in the code, or an issue in purchased products. Tests serve to bring these weak points to light before they can become an entry point for an attack.

The value of proximity and European autonomy

When an organization asks to train employees or verify networks and applications from the inside, trust is not a commercial detail. It matters to know who the people involved are, how they work, and what experience they have gained. Proximity can make the difference precisely because cybersecurity cannot be compared to just any product purchased from an unknown supplier. The comparison once again is with medicine: one would hardly entrust one’s health to someone whose skills and reliability are unknown. The same principle applies to systems, data, and digital infrastructures. This perspective also includes the issue of European digital independence. Europe has been working for decades on rules designed to protect data and limit the impact of technologies. Ignoring this path by relying entirely on purely commercial solutions, perhaps distant and difficult to control, would not be consistent with the level of attention built up over time. Digital independence, in this assessment, is an objective to pursue: it does not solve all problems by itself, but it increases responsibility and awareness in technological choices. The geopolitical tensions of recent years make this need even more evident, especially when it comes to cybersecurity. The overall picture brings together several dimensions: the acceleration of artificial intelligence, the fragility of organizations with limited resources, the need for continuous training, the technical verification of defenses, the role of ethical hackers, the need for reliable suppliers, and the prospect of greater European autonomy. For SMEs, local public administrations, and healthcare organizations, cybersecurity therefore does not appear as a goal to be achieved through a single purchase, but as a path to be built over time, by choosing proportionate measures, recognizable expertise, and relationships based on trust.