Effective Date: July 1, 2025
Version: 1.0

1. Introduction

At CliniScripts Inc. (“CliniScripts”, “we”, “our”, or “us”), we are committed to the ethical and responsible deployment of Artificial Intelligence (“AI”) technologies. This AI Safety & Ethics Policy outlines how we design, monitor, and govern our AI systems to ensure they remain aligned with clinical values, legal frameworks, and user trust.

CliniScripts provides AI-powered scribe services to clinicians, particularly in sensitive sectors such as mental health. Our goal is to enhance—not replace—clinical workflows, while upholding the highest standards of transparency, privacy, safety, and ethical responsibility.

2. Human-Centered Design

• AI is designed to support, not replace, clinical judgment.
  
  
• The clinician always retains final authority over notes, decisions, and EMR entries.
  
  
• No AI-generated content is submitted without human review.
  
  

3. Transparency & Explainability

• All AI outputs include confidence scores, highlighted reasoning, or saliency cues.
  
  
• Users are informed of the AI’s limitations and educated on when to override or disregard its suggestions.
  
  
• All AI activity is traceable and reproducible for audit, legal, or clinical review.
  
  

4. Bias Mitigation

• Our models are audited for bias related to race, gender, language, and medical history.
  
  
• We do not deploy any model that demonstrably disadvantages vulnerable populations.
  
  
• Bias monitoring is an ongoing process, not a one-time test.
  
  

5. Data Privacy & Consent

• We operate in compliance with HIPAA (US), PHIPA (Canada), and PIPEDA (Canada), depending on the user’s jurisdiction.
  
  
• AI usage is contingent upon explicit patient consent, either direct or via clinic policy.
  
  
• All sessions are end-to-end encrypted, and no patient-identifying data is stored beyond the session unless integrated into an EMR by a human clinician.
  
  

6. Safety Controls

7. Clinical AI Ethics

• Do No Harm: CliniScripts will not provide diagnosis or treatment without clinician involvement.
  
  
• Informed Use: All users are made aware of AI limitations, capabilities, and escalation protocols.
  
  
• No Surveillance Use: CliniScripts AI is never used for scoring behavior, surveillance, or law enforcement purposes.
  
  
• Reproducibility: Every AI action is traceable and reproducible for medico-legal integrity.
  
  

8. Incident Response & Kill-Switch Protocol

In the rare event of unexpected or unsafe AI behavior:

• The session is immediately terminated.
  
  
• A kill-switch disables the model inference system.
  
  
• A clinician alert is triggered, and incident logs are automatically submitted for review.
  
  
• The affected model will not be redeployed until safety retesting is completed.
  
  
• All incidents are reviewed within 24 hours by the Ethics & Safety Officer.
  
  

9. Governance Structure

• Chief Ethics & Safety Officer (CESO): Oversees audits, incident response, and the AI safety roadmap.
  
  
• AI Safety Committee: Cross-disciplinary group of clinicians, engineers, and compliance experts who meet quarterly.
  
  
• Third-Party Audits: Annual audits are conducted by independent safety and compliance reviewers.
  
  
• Whitepapers & Transparency Reports: We publish yearly safety summaries for public and clinical review.
  
  

10. Ethical & Legal Alignment

Our AI systems are designed in accordance with:

• OECD AI Principles
  
  
• EU AI Act (High-Risk Medical AI category)
  
  
• WHO Guidance on Ethics and AI for Health
  
  
• Partnership on AI – Safe Deployment Framework
  
  

11. Policy Updates

We reserve the right to update this AI Safety & Ethics Policy at any time. Any changes will be reflected on this page and, where appropriate, communicated to our users.

12. Questions or Concerns

If you have questions, concerns, or complaints about CliniScripts’ AI practices or safety protocols, please contact:

• Email: ethics@cliniscripts.ai
  
  

Response Time: Guaranteed within 72 hours