1. Our AI Mission Statement

2. Core Ethical Principles

• Human Oversight: Every high-impact AI suggestion is subject to review by qualified bioinformaticians or clinicians. Users have the right to know when an insight is AI-generated.
• Scientific Validity: Our models are trained on peer-reviewed genomic datasets and updated weekly to reflect the current state of medical science.
• Algorithmic Fairness: We actively audit our models to identify and mitigate “biogeographical bias,” ensuring our AI performs accurately for individuals of all ancestral backgrounds.
• Transparency & Explainability: We strive to provide “interpretable AI.” Where possible, our reports explain why the AI reached a specific conclusion, citing the relevant genetic markers and medical literature. We also aim to publish our AI work in high reputational and scientifically peer reviewed journals you can find some of our work at this link
• Contestability: Users and clinicians have a clear pathway to challenge an AI-generated insight if it contradicts clinical observations.

3. Data Governance & Model Training

• No Public AI: We never input your identifiable genetic or health data into public LLMs (like ChatGPT or Claude). All processing happens within our secure, private VPC (Virtual Private Cloud).
• De-identification: Only de-identified, aggregated data is used for model training and refinement. Your name and identity are stripped from the datasets used to teach our algorithms.
• Opt-out Rights: Under GDPR and Australian law, you may opt-out of having your de-identified data used for “secondary research” or model training without affecting your access to the primary service.

Compliance with High-Risk AI Standards

• Risk Management Systems: Continuous monitoring for “model drift” (where AI accuracy decreases over time).
• Data Quality: Strict “Data Hygiene” protocols to ensure the training data is representative and free from corruption.
• Technical Documentation: Detailed “Model Cards” that log the training history, versioning, and intended use of every algorithm in production.

5. Prohibited AI Practices

• Predictive Profiling for Insurance: We do not provide data or tools to insurers to predict future health costs for the purpose of denying coverage.
• Social Scoring: We do not use genetic data to rank or score individuals for social or employment opportunities.
• Unsupervised Diagnosis: Our AI is forbidden from issuing a formal medical diagnosis directly to a consumer without a healthcare professional intermediary.

6. Transparency Disclosure

Contact the AI Ethics Committee