AI Safety Information

1. INTRODUCTION

AverCare is committed to developing, deploying, and operating Artificial Intelligence (AI) systems in a safe, responsible, transparent, and clinically appropriate manner.

This AI Safety Information Notice explains the safety measures, controls, governance practices, and oversight mechanisms used within the AverCare ecosystem to help ensure that AI technologies are used safely and responsibly.

This document should be read together with:

• Privacy Policy
• Terms of Service
• AI Privacy Information Notice
• Responsible AI Framework
• AI Ethics Policy
• AI Transparency & Disclosure Policy
• Human-in-the-Loop Policy

2. OUR COMMITMENT TO AI SAFETY

AverCare is committed to ensuring that AI systems:

• Support patient wellbeing
• Promote positive health outcomes
• Operate safely and responsibly
• Respect human autonomy
• Support healthcare professionals
• Protect privacy and confidentiality
• Minimize risks
• Remain subject to human oversight

Patient safety and user wellbeing always take priority over automation, efficiency, or commercial objectives.

3. HOW AI IS USED AT AVERCARE

Artificial Intelligence may be used to support:

• Health education
• Wellness coaching
• Preventive healthcare
• Health tracking
• Wearable device analytics
• Health insights
• Clinical decision support
• Risk assessments
• Personalized wellness experiences
• User engagement

AI is intended to assist users and healthcare professionals and is not intended to replace professional medical judgment.

4. HUMAN OVERSIGHT OF AI

AverCare utilizes a Human-in-the-Loop (HITL) governance model.

This means:

• Human reviewers oversee clinically significant AI outputs
• Healthcare professionals retain decision-making authority
• AI recommendations may be modified by humans
• AI recommendations may be rejected by humans
• AI recommendations may be overridden by humans
• Human judgment always takes precedence over AI outputs

AI systems are not permitted to make autonomous clinical decisions.

5. CLINICAL SAFETY CONTROLS

AverCare implements safeguards to help reduce clinical risks.

These safeguards include:

• Clinical review processes
• Human approval workflows
• Escalation procedures
• Validation testing
• Safety monitoring
• Medical, Health & Wellness advisory oversight
• Quality assurance reviews
• Clinical governance controls

Clinically significant outputs may require review by qualified healthcare professionals before action is taken.

6. AI MODEL VALIDATION

Prior to deployment, AI systems undergo validation activities that may include:

• Accuracy testing
• Performance testing
• Clinical validation
• Safety testing
• Security testing
• Bias assessments
• User acceptance testing
• Human oversight testing

Validation activities are intended to confirm that AI systems perform according to their intended use.

7. CONTINUOUS MONITORING

AI systems are subject to ongoing monitoring.

Monitoring activities may include:

• Performance reviews
• Safety assessments
• Incident monitoring
• User feedback analysis
• Bias monitoring
• Accuracy reviews
• Model drift monitoring
• Escalation reviews

Monitoring helps identify issues that may require corrective action or additional safeguards.

8. HUMAN OVERRIDE CAPABILITY

Where appropriate, AverCare systems include mechanisms allowing authorized personnel to:

• Modify AI recommendations
• Reject AI recommendations
• Override AI recommendations
• Escalate concerns
• Suspend AI-driven workflows

Human reviewers always retain authority over AI-generated outputs.

9. AI LIMITATIONS

Artificial Intelligence systems have limitations.

AI systems may:

• Produce inaccurate outputs
• Produce incomplete outputs
• Misinterpret information
• Generate recommendations requiring clarification
• Perform differently across use cases

AI-generated outputs should not be considered infallible. Users should exercise appropriate judgment and seek professional advice where necessary.

10. EMERGENCY MEDICAL LIMITATIONS

AverCare AI systems are not intended to provide emergency medical services.

AI systems should not be relied upon during:

• Medical emergencies
• Life-threatening situations
• Acute medical crises
• Situations requiring immediate medical intervention

If you believe you are experiencing a medical emergency, contact local emergency services immediately or seek urgent medical care.

11. WEARABLE DEVICE SAFETY

Health insights generated from wearable devices may incorporate AI-assisted analysis.

Users should understand that:

• Wearable devices may not capture all health events
• Device readings may be affected by usage conditions
• Wearable insights may not constitute medical diagnoses
• Healthcare professionals should be consulted regarding medical decisions

Wearable data should be considered one source of information among many.

12. HEALTH COACHING SAFETY

AI-assisted health coaching services are intended to:

• Support education
• Encourage healthy habits
• Promote wellness
• Support user engagement

Health coaching services are not intended to:

• Diagnose conditions
• Prescribe medication
• Replace healthcare professionals
• Provide emergency medical advice

13. FAIRNESS AND BIAS MANAGEMENT

AverCare is committed to minimizing bias in AI systems.

Measures include:

• Bias assessments
• Clinical reviews
• Validation activities
• Fairness testing
• Human oversight
• Continuous monitoring

AverCare continuously evaluates AI systems to reduce the risk of unfair outcomes.

14. PRIVACY AND SECURITY SAFEGUARDS

Information processed by AI systems is protected through:

• Encryption
• Access controls
• Multi-Factor Authentication
• Monitoring and logging
• Security assessments
• Vulnerability management
• Incident response procedures
• Data governance controls

These safeguards are designed to protect personal and health-related information.

15. INCIDENT RESPONSE

AverCare maintains procedures for managing AI-related incidents.

Examples include:

• Harmful outputs
• Incorrect recommendations
• Safety concerns
• Security incidents
• Privacy incidents
• System malfunctions
• Ethical concerns

Incidents are investigated and addressed through established governance processes.

16. USER RESPONSIBILITIES

Users are encouraged to:

• Review information carefully
• Verify important information
• Consult qualified healthcare professionals
• Report concerns
• Follow medical advice from licensed professionals
• Seek emergency assistance when necessary

AI systems should be used as supportive tools and not as substitutes for professional healthcare.

17. REPORTING AI SAFETY CONCERNS

Users may report AI-related concerns, including:

• Incorrect outputs
• Potential safety issues
• Suspected bias
• Technical problems
• Security concerns
• Privacy concerns

Reports help AverCare improve safety and system performance.

18. CONTACT INFORMATION

Questions regarding AI safety may be directed to:

AI Governance Committee · ai-governance@avercare.global

Privacy Officer · privacy@avercare.global

Data Protection Officer · dpo@avercare.global

General Support · support@avercare.global

19. CHANGES TO THIS NOTICE

AverCare may update this AI Safety Information Notice from time to time.

Updates may occur due to:

• Regulatory changes
• Technology changes
• Safety improvements
• Operational enhancements
• New AI capabilities

Updated versions will be made available through AverCare platforms.

20. VERSION CONTROL

09 June 2026 Date: ______________________________ 09 June 2026

This is the published version of AC-PUB-AIS-161, version 1.0. The signed original is retained on record and is available to customers, auditors and regulators on request via ai-governance@avercare.global.