AI Safety Principles & Guardrails.
ReadInward is committed to building a safe, supportive, and privacy-respecting reflection space for personal development.
Software Reading Coach, Not Therapy
ReadInward is a software-based reading companion designed to support self-reflection. It is not professional therapy, a medical diagnostic service, a clinical intervention, or a crisis response tool. We encourage users seeking professional mental health support to connect with licensed clinical providers.
Data Minimization & Absolute Privacy
Your reflection prompts and coach conversations are strictly confidential. We do not sell your reflection data, nor do we use your personal reflections to train public LLM models. All AI responses undergo automated redaction to strip personally identifying information (PII) before logging.
Strict Safeguards & Support Rails
Every AI response surface contains reporting options to let users flag hallucinations, clinical claims, or unsafe outputs. When crisis or self-harm keywords are detected, the app automatically intercepts the conversation and presents region-appropriate, unauthenticated crisis resources.