AI-augmented psychotherapy is an approach to expanding the capabilities of psychotherapy through the deliberate integration of AI and connected digital infrastructure into clinical practice, the therapeutic process, method development, and institutional activities. It preserves the method’s theoretical foundation, human responsibility, and the central importance of the therapeutic encounter.
Psychotherapy has developed its own theories of the person, psychotherapeutic methods, professional schools, educational institutions, research traditions, and forms of supervision. Yet its technological layer has largely developed separately—as a collection of external tools, platforms, and general-purpose digital services.
Until now, psychotherapeutic methods have rarely built their own technological environments around their understanding of the person and grounded in their theory.
AI and digital infrastructure can be designed not around the abstract task of ‘helping a user,’ but from within a particular psychotherapeutic method: its language, view of the person, working principles, professional boundaries, and logic of development.
This augmentation can take place across four interrelated scales.
AI can extend capabilities beyond those of an individual tool or practitioner. The concept encompasses clinical work, the development of a psychotherapeutic method, and its wider professional environment.
AI can help psychotherapists attend to a broader range of relevant information: review a client’s history and therapy material, find relevant concepts in the method’s literature, compare definitions, revisit earlier questions, organize notes, and consider alternative hypotheses.It can help distinguish observation from interpretation and prepare material for supervision. Its role is to support professional reflection while leaving clinical decisions with the psychotherapist.
AI can support continuity between sessions by helping a person record experiences, connect fragments of experience, return to their own wording, and prepare meaningful material for the next session.Such an environment cannot experience or work through personal difficulties on someone’s behalf, nor can it replace the relationship with a psychotherapist. It can offer temporary support for reflection and psychological work.
A method’s theory can exist as coherent, versioned knowledge that preserves the history of changes, the origins of concepts, their relationships, internal debates, and competing lines of development.AI can help identify conceptual connections and contradictions, compare formulations, and bring questions to the clinicians and researchers responsible for developing the method. This preserves a record of both the method’s current position and how it developed.
Knowledge from psychotherapeutic practice, supervision, education, research, and theory development is often held in separate professional domains. Important connections between them often remain invisible to any one person.A shared professional memory allows AI to locate and compare elements of this knowledge while preserving their origins and historical context. A question from practice can be connected to theory and research; research findings can inform education, supervision, or a new version of a tool; practitioners can see how the method has changed.AI can help the professional community identify relationships within a body of knowledge too large for any one person to keep in mind. Their interpretation and evaluation remain professional responsibilities.
AI-augmented psychotherapy is a universal concept. It does not belong to a single psychotherapeutic approach and does not prescribe one technological model for every method.
It can be expressed as Xai, where X is a specific psychotherapeutic method and the suffix ai denotes its AI-augmented form.
This makes systems such as TFPai, DBTai, and others possible, each grounded in its own theory, professional language, and view of the person.
The universal concept becomes concrete only when a psychotherapeutic method creates a technological environment appropriate to it.
Transference-Focused Psychotherapy (TFP) and Dialectical Behavior Therapy (DBT) name established methods. TFPai and DBTai are used here only as model labels for the Xai pattern, not as names of existing products or ecosystems. DET → DETai is an emerging example of this pattern.
a specific method × its AI-augmented form
For comparison with the original descriptions of these methods, see the following authoritative sources: TFP / International Society of Transference-Focused Psychotherapy · DBT / Linehan Institute.
Method-oriented AI infrastructure is a specialized, dynamically evolving digital environment that holds and operationalizes the theory, professional language, and practical boundaries of a specific psychotherapeutic method.
Such infrastructure does more than provide practitioners with ready-made knowledge and tools. It connects method theory, technological solutions, psychotherapeutic practice, education, and research in a managed development cycle.
An observation or difficulty arising in one domain can become a question for another. But a connection identified by AI changes nothing by itself: it must pass through professional interpretation, verification, and a deliberate decision by the people responsible for developing the method.
Each proposed change has a documented origin, undergoes professional review, and receives an explicit decision, which may range from implementation to the archiving of a hypothesis.
1
The theory remains unchanged, while an interface, tool, AI agent, or way of presenting knowledge is updated.
2
Experience shows that an existing proposition of the method needs a more precise practical explanation or distinction.
3
Accumulated questions and verification results provide grounds for revising a concept, proposition, or relationship within the method’s theory.
4
One change simultaneously affects theory, a technological tool, education, a research protocol, or psychotherapeutic practice.
5
Only the substantiated part of a proposed change is accepted; the remainder stays open for further inquiry.
6
A hypothesis is not yet ready for inclusion in the method’s established body of knowledge, but remains available for further development and testing.
7
A change may be rejected, retained in the development history, or returned for revision with the reasons for the decision recorded.
DET is an independent psychotherapeutic method with its own theory of the person, professional language, and principles of therapeutic practice.
DETai is its AI-augmented form and the ecosystem developing around it. It is at an early stage of development and aims to put AI-augmented psychotherapy into practice, from supporting individual practitioners to building a technological environment that can contribute to the development of the method itself.
Within this ecosystem, DET theory provides the basis for designing AI agents, digital tools, and professional infrastructure. The method’s understanding of the person guides technological development, while questions arising from that work are brought back to the professional community for discussion and investigation.
DETai’s method-oriented AI infrastructure is being developed to put this relationship into practice. Its structure and interaction with the ecosystem’s other components are explained in the Platform DETai section.
