The 'human' must be the pivot of digital regulation

The European Ai Act and related regulations represent an unprecedented, perhaps hypertrophic, effort of technological governance, which aims to build a framework of rules capable of combining technological development and the protection of fundamental rights. Of course, good intentions do not always lead to desired results.

The Ai Act also introduces fundamental principles that must guide the design, development and deployment of Ai: these include technical robustness, security, privacy protection, and respect for human rights. All this is of added value if the European regulator, in collaboration with the competent national authorities, will find a way to concretise and even earlier make these principles comprehensible for the Ai Act's governance and implementation season. Otherwise the risk is not only that of a missed step forward, but of an irreversible step backwards in the search for a sustainable balance between innovation and regulation

The new European legislation in the light of the above, and provided the conditions mentioned before are realised, can represent a crucial step in the direction, mentioned before, of the implementation of a multilevel model. With the risk-based approach, Europe has chosen to set high standards for high-impact systems, imposing validation, explainability and human oversight. This is an important signal: confidence in innovation depends on ensuring that digital tools do not become an uncontrollable 'black box'.

Come to think of it, this applies to all areas where prevention is measured by technology. Gender violence is an example: data can help intercept recidivism or report critical situations, but support networks, human professionalism and targeted interventions are needed. The fight against racism requires tools to monitor hate online, but also cultural work that cannot be automated. Terrorism and digital radicalisation pose the challenge of balancing security and freedom: the algorithm can cross digital traces, but not replace the control of legality.

Precisely for this reason, each area deserves a vertical analysis. It will be the task of the next contributions to explore, from time to time, how to strike this balance between prevention, technology and rights. A general premise applies here, however: prevention is not just a question of data and software, but of institutional culture. Without training, without widespread responsibility, without awareness of limits, no algorithmic tool will be able to compensate for the fragilities of a system.

Those who feed the illusion of an all-powerful technology, capable of replacing human relations and constitutional headmasters, risk feeding dangerous misinformation. The real challenge is to show that regulating does not mean blocking innovation, but making it sustainable. Governing technology means ensuring that it remains an instrument of freedom, not control.

At a time when risks change shape with impressive speed, prevention cannot remain stuck in patterns of the past. But neither can it be reduced to an automatic calculation. It requires vision, culture, guarantees. And, above all, it requires remaining faithful to a non-negotiable principle: even in the season of the algorithm and the autonomy of Ai, the person must remain at the centre. And those who separate efficacy and humanity in the design of the prevention measures evoked earlier tell of a dichotomy that does not exist. A dangerous dichotomy, to be firmly opposed.