AI Governance

Last updated: February 2026

This page describes how we manage AI behavior, safety guardrails, user feedback, and updates. Enhanced Genetics AI provides informational insights and is not a medical provider.

1) Governance objectives

  • Safety-first: prioritize user safety and non-diagnostic outputs.
  • Transparency: publish clear explanations of AI behavior and limitations.
  • Accountability: maintain channels for reporting concerns and correcting issues.
  • Continuous improvement: update guardrails and policies as we learn.

2) AI guardrails

  • Outputs are designed to be informational rather than diagnostic.
  • We avoid definitive medical conclusions when data is incomplete or ambiguous.
  • We encourage clinician review for concerning results or symptoms.

See AI Safety for responsible-use guidance and escalation paths, and AI Methodology for how outputs are generated.

3) Feedback and issue handling

If a user reports an unsafe, misleading, or inappropriate AI output, we may:

  • Review the input context and generated output (as available).
  • Determine whether the issue relates to data extraction, missing context, or AI behavior.
  • Adjust prompts/guardrails or extraction logic where appropriate.
  • Document significant user-facing changes in the AI Changelog.

Safety contact:  info@enhancedgeneticsai.com

4) Update management (change control)

  • Planned updates: improvements to summaries, extraction accuracy, and UX.
  • Safety updates: guardrail and policy updates intended to reduce risk.
  • Rollback: if an update causes harmful or confusing outputs, we prioritize rollback or remediation.

Major user-facing changes are recorded in the AI Changelog.

5) Data and privacy commitments

  • We do not sell personal information.
  • Subscriptions are processed through Stripe; we do not store raw payment card numbers.
  • We apply safeguards designed to protect sensitive data.

See Privacy Policy and Security & Compliance. For HIPAA-specific information (where applicable), see HIPAA Notice.

For transparency around what data is used to generate AI outputs, see AI Data Sources.

6) Limitations

AI may be inaccurate, incomplete, or require clinical context. Read AI Limitations.

Related AI Pages