AI Features Terms of Use

PsychoSpace provides the following AI-assisted features: a) Session note summaries — automatic generation of condensed summaries of therapist's clinical notes; b) SOAP formatting — conversion of notes to standard SOAP format (Subjective, Objective, Assessment, Plan); c) Theme extraction — identification of key themes discussed during sessions; d) Note polishing — stylistic and grammatical correction of clinical notes; e) Matchmaking chat — AI assistant helping patients in the therapist matching process; f) Match reason generator — AI generating explanations for why a particular therapist was suggested. All AI features are exclusively supportive and informational in nature. They do not constitute medical advice, diagnosis, or therapeutic recommendation.

AI features are delivered through the following providers: a) OpenAI (GPT-4o) — note summaries, SOAP formatting, theme extraction, note polishing; b) xAI (Grok) — matchmaking chat, match reason generator. Both providers are headquartered in the United States. Data transfer outside the European Economic Area is conducted based on Standard Contractual Clauses (SCC).

PsychoSpace applies multi-layered data protections for data sent to AI models: a) Data sanitization (PII removal) — before sending to the AI model, data is automatically cleaned of personal identifiers (names, surnames, national ID numbers, email addresses, phone numbers, addresses); b) Prompt injection guard — a security system that detects and blocks attempts to manipulate the AI model through malicious instruction injection; c) E2EE health data — the content of emotion journals and clinical notes encrypted end-to-end is NOT sent to AI models in decrypted form. AI features operate exclusively on data that the therapist consciously submits for processing; d) No storage — PsychoSpace does not store data sent to AI models or model responses beyond the context of the current user session. Note: Despite sanitization, it cannot be excluded that in rare cases, data fragments may be processed by the AI model. Use of AI features is voluntary.

The user understands and accepts that: a) AI features may generate inaccurate, incomplete, or inadequate content — so-called AI model "hallucinations"; b) Results generated by AI do NOT constitute medical advice, diagnosis, or therapeutic recommendation; c) The therapist bears sole responsibility for the content and accuracy of clinical notes, regardless of whether they used AI features; d) Patients should not make health decisions solely based on AI-generated information; e) PsychoSpace does not guarantee the availability, accuracy, or continuity of AI features; f) AI features may be temporarily unavailable due to provider outages (OpenAI, xAI), maintenance work, or changes in provider APIs.

Therapists using AI features undertake to: a) Verify all AI-generated content before using it in clinical documentation; b) Not submit data enabling direct patient identification (name, surname, national ID) to AI features, if sanitization has not automatically removed such data; c) Not use AI results as the sole basis for diagnostic or therapeutic decisions; d) Inform patients about the use of AI tools in the therapeutic process, if required by professional ethical standards; e) Promptly report to the Platform any cases of AI generating harmful, inappropriate, or patient-rights-violating content.

Patients using the PsychoSpace platform should know that: a) The matchmaking chat uses AI to assist in the therapist matching process — AI responses are informational and do not constitute medical advice; b) The therapist may use AI features to assist in maintaining clinical documentation — the final content of notes is always verified and approved by the therapist; c) Data from the matchmaking questionnaire may be processed by AI models to generate recommendations — data is previously sanitized (PII removed); d) The therapist's use of AI features is optional and does not affect the quality of therapeutic services provided.

The therapist may disable all AI features at any time in their profile settings. Disabling AI features: a) Does not affect the availability of other Platform features; b) Does not cause loss of data or clinical notes; c) Can be reversed at any time. Patients cannot disable AI features used by the therapist, but may request the therapist not to use AI tools in the context of their therapy.

PsychoSpace reserves the right to: a) Add new AI features; b) Modify existing AI features; c) Change AI service providers; d) Temporarily or permanently disable AI features. Users will be informed of significant changes to AI features with 14 days' notice via in-app notification or email.

For questions about AI features, please contact: Data Protection Officer: iod@psychospace.io PsychoSpace sp. z o.o. ul. Cyfrowa 6, 71-441 Szczecin Reports about AI feature malfunctions: support@psychospace.io