Artificial Intelligence and Psychic Care
Artificial intelligence does not burst into the field of psychic care with fanfare. It enters quietly, through everyday uses, mobile interfaces, conversational tools. It becomes a presence without a body, a voice without breath, knowledge without gaze. This shift, discreet yet decisive, questions the very nature of therapeutic address.
Traditionally, psychic care is rooted in shared speech, in the construction of a slow, embodied relational space made of waiting, transference, and silent welcomes. The digital device, by contrast, offers immediacy, constant availability, algorithmic response. Can we still speak of a “meeting” when the interlocutor neither doubts, nor withdraws, nor dreams?
Yet AI undeniably brings benefits. It offers support to people in distress, sometimes at hours when no clinician is reachable. It enables personalized, adaptive follow‑up based on physiological and behavioral data. It identifies early vulnerabilities and directs patients toward useful resources. In certain contexts, it acts as a lever for inclusion, reducing geographic and social distances.
But this potential is accompanied by a series of ethical and clinical concerns. Psychic data are among the most sensitive: what becomes of their confidentiality when they pass through servers, clouds, probabilistic models? Algorithms are trained on historical corpora, often biased — which subjectivities risk being misinterpreted, under‑diagnosed, or rendered invisible?
Take the example of Clara, 31, prone to anxious‑depressive episodes. A conversational agent accompanies her daily: it suggests breathing exercises, detects mood fluctuations, encourages her to open up. Clara describes a form of relief… but also a strange sensation of amplified solitude. She speaks of a “bodiless dialogue,” of a listening that returns neither gaze, nor otherness, nor tremor.
What we call “care” cannot be reduced to efficiency. It involves a subjective encounter, a crossing of words and silences. It presupposes the existence of another — vulnerable, fallible — capable of welcoming without resolving, containing without correcting. However sophisticated, artificial intelligence cannot offer that. It can accompany, support, prevent, but it cannot embody the human presence that grounds the clinical frame.
Thus, far from opposing humanity and technology, we must think their articulation. AI can become a valuable ally, provided it never replaces the clinician’s singular listening. Care, ultimately, remains a space of otherness and shared temporality — a place where something happens, not through treatment, but through connection.
