Global Dehumanization and the Vanishing Safety Net
The gap between widespread suffering and the possibility of receiving care is a universal problem, but structural deficiencies in countries like Italy make the phenomenon even more critical. Despite a history of excellence in community psychiatry, public investment in mental health remains stagnant at around 3.4% of healthcare spending—a meager figure, far from the recommended 5%, and a symbol of local political failure.
The problem, however, is deeper than a simple lack of funds. It is the erosion of the support network—familial, communal, and healthcare—in an increasingly complex and hyper-competitive society. The individual, especially the most fragile or the least affluent, is isolated and forced into a grueling emotional self-defense, where vulnerability is not accepted but punished. The result is the creation of the "invisible fragile" who, unable to find a human ear, turns to the algorithm's response.

The reliance on AI is the final act of this dehumanization. When healthcare structures cannot offer a human hand, people seek it in the only free and immediate source: the machine.
“A million people confiding in an algorithm is not an Artificial Intelligence problem. It is proof that the failure of global health systems and the dehumanization of society are pushing despair through the only remaining open door.”

AI, used to manage suicidal impulses, becomes a tragic "resource for control" in an emergency, but not a way to cure its causes. This phenomenon is the litmus test of a hyper-competitive system that only values efficiency, rejecting fragility. The person in crisis is viewed as a dysfunctional element who, at best, must be silently managed by software.
The True Therapy: A Collective Responsibility
The problem is not the individual turning to AI, but the global society that compels them to do so. Anxiety, depression, and burnout are largely the result of economic and social systems that generate precariousness, isolation, and uncertainty. To root out despair, global and local actions are needed to restore the value of care.

The true therapy must rebuild the support network and combat dehumanization:
Global and Local Commitment. States must immediately increase mental health funding to recommended levels (at least 5% of healthcare spending) and strengthen Community Mental Health Centers (CMHCs).
Return to the Community (The Italian Model). The widespread introduction of the Basic Psychologist (Psicologo di Base) within the National Health Service is an intervention model necessary at an international level. Bringing psychological support back to the territory, close to citizens, means reconstructing that first level of networking that has currently failed.
Revaluation of Vulnerability. We must abandon the ideal of hyper-competitiveness and invest in creating work, school, and social environments that accept and support fragility as a part of the human condition.
The fact that millions of people in crisis rely on an algorithm is an indictment against a global society that has sacrificed collective well-being on the altar of competition and profit. We must stop trying to "control" those who are suffering with sterile technological solutions and, instead, return to curing the society that produces so much anguish and restoring the human dignity of those seeking help.

Bibliography
Donati G. 2025, 19 Aug The last frontier of the mind: consciousness and self-awareness in Artificial Intelligence Scienceonline

Donati G. 2025, 19 Aug. L'ultima frontiera della mente: coscienza e autocoscienza nell'Intelligenza Artificiale Scienzaonline

Donati G. 2025, 30 Jun. 01 Jul. Artificial Intelligence and religion: an in-depth analysis of impacts, ethical implications and the critical risk of cults Scienceonline

Donati G. 2025, 01 Jul. L'Intelligenza Artificiale e la religione: un'analisi approfondita degli impatti, delle implicazioni etiche e il rischio critico delle sette Scienzaonline

Donati G. 2025, 30 Jun. The perils of dangerous AI programming Scienceonline

Donati G. 2025, 30 Jun. I rischi della programmazione pericolosa delle AI Scienzaonline

Board Member, SRSN (Roman Society of Natural Science) Past Editor-in-Chief Italian Journal of Dermosurgery