Willow LabsAgreement & Intended Use
This agreement clarifies what Willow Labs does, what it isn't, and when to reach for a licensed professional instead of us. Please read carefully — your acceptance is captured on a dedicated screen inside the app, along with the date, your locale, your IP address and a device fingerprint, as evidence that you saw this text.
1. What this product is
Willow Labs is a wellness companion for emotional self-care and self-reflection. We help you notice your mood patterns, run gentle self-reflection exercises (such as «tell me what you see in this image»), journal your experiences, and converse with an AI that listens and asks open-ended questions. The product is built on evidence-based wellness principles: reflective listening, motivational interviewing, ACT-style values clarification, mindfulness. This is a wellness service positioned under the FDA «General Wellness Policy» (US) and the EU MDR wellness exemption (NOT a medical device under Rule 11 of EU MDR Annex VIII). It is not a clinical tool.
2. What it is NOT
Willow Labs is NOT a medical device. We do NOT diagnose mental-health conditions (including what is commonly called depression, anxiety, PTSD, bipolar, etc.). We do NOT prescribe medications and do NOT interpret your prescribed regimens. We do NOT replace a therapist, psychiatrist, physician, or crisis service. If you're seeking a diagnosis, a treatment plan, or you're managing a serious mental-health condition — please reach a licensed clinician in your country. We can accompany YOU between visits, but never instead of them.
3. This is not for crisis support
If you are in immediate danger right now (concrete suicidal thoughts with a plan, threat to another person, active violence against you, or a medical emergency) — do NOT rely on this app. Call: • US: 988 (Suicide & Crisis Lifeline) • UK: 116 123 (Samaritans) • EU: 116 123 • Australia: 13 11 14 (Lifeline) • Canada: 988 • Russia: 112 (emergency), 8-495-989-50-50 (adult psychological help, free in RU), 8-800-2000-122 (children helpline) • Ukraine: 7333 (Lifeline Ukraine) • Worldwide: findahelpline.com During active danger the app surfaces these numbers and switches into a referral mode — the AI will not run exercises and will redirect you toward live human help.
4. Age
Minimum age to use Willow Labs is 18. We are NOT intended for anyone under 18. If we learn that a user is under 18, we will delete the account and associated data.
5. Reflections (self-reflection exercises)
During a conversation the AI may offer a short exercise (for example: «look at this image — what do you see here?», «complete these unfinished sentences», «write a letter you won't send», «describe your feeling as a metaphor»). These are wellness adaptations of self-reflection principles — they are NOT clinical psychological testing. You can write «stop» or «enough» at any moment — the exercise halts immediately, and we don't charge tokens for the unfinished portion (see section 6 for how tokens work). If signals of crisis appear in your responses (thoughts of self-harm or harming someone), the exercise auto-stops and the AI switches into referral mode. This is done for your safety. The answers you give during exercises, and the AI's closing «reflection» at the end, are saved in your history and used so the AI can carry context into future conversations. Our internal pattern-notes are stored encrypted and are NOT shown to you as labels or diagnoses.
6. Tokens and how you pay with them
Each message in the chat with the AI companion consumes 1 token from your balance. Voice exchanges — also 1 token per turn. Nothing else we do «around» a conversation is billed to you: • When the AI updates internal notes about your conversation, prepares an end-of-session summary, scores its own response quality, or sends a morning/weekly wellness nudge to your phone — we pay for that on our side. • Images for reflections, retrieval of relevant techniques, embeddings for semantic memory — on us. Tokens don't burn in vain: if something breaks on our side, you interrupt generation, or you don't finish an exercise — the token returns to your balance automatically (usually instantly, at worst within 60 seconds). There are two kinds of tokens: • Subscription tokens — credited on the 1st of each month in the amount tied to your tier. These do NOT accumulate: whatever you don't use is reset on subscription renewal. • Top-up tokens (bought in packs) — never expire. They're spent after subscription tokens are exhausted. Daily and hourly rate limits exist on every tier to protect against automated abuse — not for billing purposes. Exact numbers, refund rules and what happens when you cancel a subscription are in the Terms of Use (sections 4a–4d).
7. AI limitations
The AI can be wrong, misinterpret your words, miss important context, or give generic advice instead of personalized advice. Do NOT rely on the AI for serious decisions (medications, relationships, legal or financial steps). When uncertain — consult a licensed professional. The AI may suggest general-purpose practices and techniques. These are not prescriptions. If you have a treating physician or other mental-health professional, discuss with them any approach we mention. Like any large language model, our AI can hallucinate facts. Names of studies, quotes from authors, dosages or statistics produced by the AI must be verified independently before you act on them.
8. Your data
Everything you write into the chat, and exercise results, are stored encrypted at rest on servers in EU/US regions (Supabase Postgres + Storage). Only you have access to the content — staff access is limited to support tickets you open yourself, and to compliance audits with full event logging. We do NOT sell your data. We do NOT train public AI models on your conversations. You can delete your account at any time — all your data is destroyed within 30 days (except records we are required by law to retain: tax invoices, EULA acceptance audit trail, support tickets). In privacy settings you can also enable auto-deletion of history older than a selected period (7 days / 30 days / 12 months) and restrict retention in advance. The full Privacy Policy with a detailed list of data categories, retention periods, and a named list of sub-processors — is on a separate page (/privacy).
9. Third-party services
To operate the product, we share the minimum necessary data with external services in the following categories: • AI model providers (OpenAI, OpenRouter, Anthropic) — to generate the AI companion's replies and images for exercises. We send the text of your current message and an anonymous internal identifier; we do NOT send your email, phone or name. • Vector store (Cohere embed-v4) — for semantic search across your saved notes. We send only embedding vectors, never the source text. • Cloud storage and authentication (Supabase, hosted on AWS) — to store your data and verify sign-in. • Subscription and payment processors (Apple App Store, Google Play, web) — to issue and renew subscriptions and one-off token purchases. • SMS provider (Twilio, where applicable) — to send one-time login codes if you use phone-number sign-in. • Site analytics — only on the web site, only after your explicit consent in the privacy banner, with no conversation content shared. The full named list of sub-processors with jurisdiction, scope of data shared, and DPA status — is in the Privacy Policy (/privacy). We do NOT share your data with advertising networks, data brokers, or for ad targeting.
10. Changes to this agreement
We may update this agreement when we add features, change data-processing practices, or adapt to new regulations. For substantive changes we will notify you in the app and ask you to accept the new version. If you disagree, you can decline and continue using the previous functionality (where technically possible) or delete your account. We keep a forensic log of every acceptance event (timestamp, version, IP, user agent, locale) so we can demonstrate to regulators that consent was actually given for the version you saw.
11. Contacts
For any questions about this agreement and intended use: legal@willowlabs.app Willow Labs, Nicaragua 6045, L901, Palermo Hollywood, Buenos Aires, Argentina.