Safety & Ethics

Our commitment to safety.

MindBridge is designed with clinical responsibility and patient safety at its core. We are transparent about what we do, what we don't do, and the safeguards we have in place.

What MindBridge Is

  • An assistive triage tool that helps clinicians prioritize patients based on risk assessment
  • A structured intake system using validated instruments (PHQ-9, GAD-7)
  • A support for clinicians to reduce administrative burden and improve clinical efficiency
  • Human-in-the-loop — All high-risk cases are reviewed by licensed clinicians

What MindBridge Is NOT

  • Not a diagnostic tool — We do not diagnose mental health conditions. Only licensed clinicians can diagnose.
  • Not a crisis service — If you are in immediate danger, call emergency services (911/000) or a crisis line.
  • Not a replacement for clinical judgment — AI assists, but humans decide treatment.
  • Not therapy or treatment — We collect information; we don't provide ongoing therapeutic care.

Known limitations of AI

We are transparent about the limitations and risks of AI in healthcare.

Hallucination Risk

AI can generate plausible-sounding but incorrect information. All outputs are reviewed by clinicians before clinical action.

Bias & Fairness

AI models can reflect biases in training data. We continuously monitor for demographic disparities in risk stratification.

Context Limitations

AI lacks full clinical context (patient history, physical exam, cultural nuances). This is why human review is mandatory.

Not a Black Box

We provide clinicians with the reasoning behind risk scores, key factors identified, and source data for transparency.

Safety Measures We Take

Clinical Guardrails

AI is programmed to never diagnose, never guarantee safety, and always escalate high-risk cases to human clinicians immediately.

Continuous Monitoring

We monitor all sessions for safety events, log critical incidents, and conduct regular audits of risk stratification accuracy.

Privacy & Security

We are committed to HIPAA (US) and Australian Privacy Act compliance:

  • • End-to-end encryption of patient data
  • • PHI redaction in logs and analytics
  • • Minimum necessary data collection
  • • Regular security audits and penetration testing

Validated Instruments

We use evidence-based, clinically validated assessment tools (PHQ-9 for depression, GAD-7 for anxiety) rather than relying solely on AI interpretation.

Crisis Resources

If you are in immediate danger or thinking of harming yourself, contact emergency services or your local crisis line immediately.

United States

988 Suicide & Crisis Lifeline
988
Available 24/7 • Call or Text

Australia

Lifeline
13 11 14
Available 24/7 • Call or Text

Emergency: 911 (US) or 000 (Australia)

Questions about safety?

We are committed to transparency. If you have concerns about safety, privacy, or ethics, please reach out to our clinical team.