AI Transparency
How intelligence is derived from your health data, and the nature of AI-generated outputs.
Current AI transparency
This page describes how AI-assisted features currently operate in the ARQENA informational service and may be updated as the service evolves.
Three Layers of Information
ARQENA distinguishes between three types of information you may encounter within the platform:
- Raw Data: Values directly recorded or imported from your devices and laboratory reports — steps, heart rate readings, biomarker results. These are presented as-measured.
- Derived Signals: Normalized or calculated information, such as trends over time, rolling averages, or context-adjusted values. These are computed from raw data.
- AI-Generated Interpretation: Contextual synthesis produced using AI systems — summaries, pattern observations and reasoning context. These are probabilistic outputs and should be understood as informational context, not clinical conclusions.
The Intelligence Pipeline
ARQENA uses a multi-stage reasoning process to transform raw signals into understanding:
- Acquisition: Secure intake of lab PDFs, wearable streams, and vitals.
- Validation: Verifying data provenance and source integrity.
- Normalization: Standardizing disparate units and biological references.
- Contextualization: Mapping data against your available health history.
- Correlation: Identifying relationships between multiple sources.
- Interpretation: Producing an AI-assisted, evidence-informed synthesis of biological trends.
Probabilistic Nature of AI Outputs
AI-generated outputs are probabilistic and may vary as available data, context, and model behavior change. The same data may produce different interpretations at different times. ARQENA does not treat generated interpretations as definitive medical conclusions, and neither should you.
We explicitly highlight uncertainty when data is missing, signals are conflicting, or coverage is insufficient. Human medical judgment should always take precedence over AI-generated outputs.
AI Model Providers
ARQENA may use third-party AI model providers as part of its intelligence pipeline. Specific providers and applicable data processing terms will be documented in this policy. We do not use your biological data to train public foundation models without explicit consent.
What AI Does Not Do
ARQENA's AI systems do not diagnose, prescribe, or make medical recommendations. They do not replace clinical judgment. They do not make legally binding determinations about your health status. For more information, please review our medical disclaimer.
Last reviewed: August 2026. Applicable regulatory requirements are reviewed as the service evolves.