AI in mental health care: towards a governance-driven infrastructure.

The central thesis:

Sustainable AI development in Dutch mental health care (GGZ) is only possible when technology is organised so that it is clinically embedded, governance-driven, context-sensitive, applicable across disciplines and supportive of human treatment practice.

The central shift: from AI tool to mental health care infrastructure

Many current AI applications focus on individual functions: transcription, summarisation, document generation, administrative support. Valuable — but a first step. Mental health care calls for support of processes that develop within continuity of treatment, multidisciplinary collaboration, diagnostic development and professional reflection.

The question is not only which tasks AI can automate, but how AI can strengthen the infrastructure within which mental health care is organised.

Continuity of treatment as the central principle

Psychological treatment usually develops not within individual contacts, but over long courses of treatment. In traditional care structures, treatment history often remains fragmented across records, disciplines, handovers and organisational boundaries.

AI can contribute to better access to treatment history, stronger multidisciplinary coordination and greater longitudinal coherence. The real added value lies in strengthening the conditions under which continuity is organised.

Multidisciplinary collaboration

Future mental health care will continue to depend on effective collaboration between psychiatrists, psychologists, psychotherapists, nurses, and systemic and creative therapists. AI can function as a supporting infrastructure that facilitates collaboration without replacing it — through accessible information, workflow integration and longitudinal treatment history.
Responsibility for interpretation and decision-making remains human.

Governance as a structural principle

In many sectors, governance is added after innovation has taken place. In mental health care that order falls short: the way AI is integrated into treatment processes directly affects professional responsibility, the therapeutic relationship and institutional reliability.

• Human authorisation and professional responsibility

• Transparency and auditability

• Risk management and quality assurance

• Multidisciplinary coordination as a structural condition

Scientific validation and implementation science

Sustainable AI development requires not only technological innovation but also a scientific basis. Practice-based validation and implementation science play a central role: AI systems are not only validated technically, but studied for their effect on continuity of treatment, collaboration, workflow integration, professional acceptance and organisational implementation.

The future organisation of mental health care

As AI becomes integrated into continuity, governance and collaboration, the organisation of care itself also changes — from separate systems and fragmented information environments towards integrated ecosystems. Within this development, AI shifts from digital support to integrated care infrastructure.

The future of AI in mental health care lies not in autonomous diagnostic systems, but in governance-driven, practice-based and clinically embedded infrastructures.

Summary of the executive summary · Notal Care. The full position paper contains Part I (conceptual) and Part II (workflow, implementation and infrastructure architecture).