Safety Standards & Ethical Guidelines

1. Evidence-Based Framework

Mira is developed using methodologies from Cognitive Behavioral Therapy (CBT), Gestalt therapy, and depth reflection protocols. Conversations are structured to assist users in identifying cognitive distortions, examining automatic thoughts, and building constructive behavioral habits.

2. Medical & Psychiatric Boundaries

Mira is explicitly an AI educational and self-reflection companion, NOT a substitute for a licensed psychiatrist, clinical psychologist, or medical doctor.

• Mira does not prescribe psychiatric medications or adjust prescription dosages.
• Mira does not issue clinical diagnoses (such as DSM-5 or ICD-11 classifications).
• Mira does not treat severe psychiatric conditions requiring hospitalization.

3. Crisis Intervention Protocols

Our neural system is tuned with safety guardrails that monitor for indications of acute crisis, severe self-harm, or threats of violence. In such circumstances, Mira immediately redirects the conversation, provides localized emergency contact numbers, and encourages reaching out to human professionals.

4. Zero-Knowledge Confidentiality

Complete emotional honesty requires complete privacy. We ensure that:

• Session conversations are encrypted with AES-256 at rest and TLS 1.3 in transit.
• User emails are masked into anonymous IDs for administrative safety.
• No conversation data is ever sold or utilized for third-party advertising.