ReLive.Genomics
— ethics · a living document —

Our Principles.

We work at the intersection of artificial intelligence and human biology — one of the most consequential scientific frontiers of our time. These are not guidelines. They are the constraints we build inside.

Building AI systems that learn from human biological data carries a responsibility unlike any other technology company.

The data we work with — brain signals, genomic sequences, physiological patterns — is not information about people. It is people. That distinction changes everything about how we operate.

— six principles —

The constraints we build inside.

I.principle

Biological data is user-governed. Permanently.

Every system we build — today and in the future — is architected around a single constraint: no training data moves forward without explicit, informed, revocable consent. Our workflow is staged: local preprocessing first, user review and filtering controls second, and only approved summaries plus selected data entering ReLive training systems. The individual remains the owner of their biological signature, and that ownership does not transfer when they become a customer of ReLive Genomics.

  • We will never sell, license, or monetize individual biological data.
  • We will not train models on data that users have not explicitly approved.
  • We will always provide a full export and deletion pathway on request.
II.principle

Consent is not a checkbox. It is an ongoing relationship.

Before any data collection begins, users go through a structured informed consent process — not a terms-of-service agreement designed to be skipped. We explain precisely what is collected, how it is processed, who may access it, and what happens to it after death. Consent is granular: users choose which data layers to enable. Consent is revocable: any individual can withdraw consent and delete their data at any time, entirely, with no friction. We treat consent as an ongoing relationship, not a one-time transaction.

  • Consent is required before any collection begins, no exceptions.
  • Consent is granular — users enable individual data layers independently.
  • Consent is revocable — withdrawal and full deletion are always available.
III.principle

We will tell you exactly what we do not know.

We are working at the frontier of computational biology and AI — a domain where scientific certainty is limited and ethical frameworks are still forming. We will not overstate the capabilities of our systems, the accuracy of our models, or the completeness of our understanding. When we are uncertain, we will say so. When our systems produce errors, we will disclose them. We publish our research openly so that the scientific community can scrutinize, challenge, and improve our methods. Transparency is not optional — it is how responsible science operates.

  • We publish our core research methodology openly.
  • We will not overstate the fidelity or accuracy of any AI persona system.
  • We will disclose limitations, failure modes, and uncertainty to users directly.
IV.principle

Our goal is to make humans better — not to replace them.

ReLive Genomics is building toward human-machine convergence. We believe this is one of the most important scientific and philosophical frontiers of this century. But convergence does not mean subordination of the human. Every product we build must demonstrably improve human life — preserving memory, supporting health, extending capability, or deepening understanding. We will not build systems designed to deceive humans, substitute human relationships, or exploit psychological vulnerability. The line between AI and human may blur. It will not disappear.

  • We will not build systems designed to simulate human presence without disclosure.
  • We will not exploit grief, loneliness, or vulnerability as a growth mechanism.
  • Every product we release must satisfy the question: does this make human life better?
V.principle

Safety scales with capability. We will slow down when necessary.

As our systems become more capable — as the Personal Biological Embedding deepens, as AI personas become more convincing, as our understanding of the human biological signature grows — our ethical review processes grow with them. We will not ship capabilities we do not fully understand the consequences of. We will conduct independent safety reviews before major releases. We will consult with bioethicists, grief psychologists, legal scholars, and domain experts as part of our standard development process. The principle is simple: the more powerful the capability, the more rigorous the review.

  • Safety review is a required gate before any major capability release.
  • External expertise — bioethics, psychology, law — is part of our standard process.
  • We will delay release if ethical uncertainty is unresolved. Revenue does not override this.
VI.principle

We are building for decades. Not quarters.

The people who use ReLive Genomics are trusting us with something that matters across generations — their biological legacy, their family's memories, their most personal data. That trust cannot be conditional on the company's financial performance. We have therefore made specific architectural and legal commitments to ensure that users retain access to their data and their systems regardless of what happens to ReLive Genomics as a business. Our core inference engine is open source. Users can export everything. A company shutdown triggers a Local Exit Kit — not data loss.

  • Core engine is and will remain open source — community-maintained indefinitely.
  • Company shutdown triggers a Local Exit Kit — users keep full local access forever.
  • No feature or service creates permanent dependency on ReLive Genomics infrastructure.
— at a glance —

What we will and will not do.

Sell user biological data
Selling, licensing, or sharing individual biological data with any third party for any purpose.
Never
Use unreviewed raw data for training
Sending unreviewed raw personal streams into model training without local preprocessing and user approval.
Never
Collect data without explicit consent
Any data collection without a completed, granular informed consent process.
Never
Publish research openly
All core methodology and architectural research published and accessible to the scientific community.
Always
Provide full data export and deletion
Users can export or permanently delete all their data at any time, with no friction or delay.
Always
Conduct independent safety review
External ethical and safety review before any major new capability is released to users.
Always
Maintain open source core engine
The inference engine that runs user systems remains open source and community-maintainable indefinitely.
Always
Exploit psychological vulnerability
Designing any product mechanism to leverage grief, loneliness, or emotional distress for engagement or retention.
Never
— a living document —

This document will change. We will tell you when and why.

ReLive Genomics is an early-stage company working on problems that do not yet have settled answers. Our ethical understanding will deepen as our technology matures, as we engage with users, as we consult with experts, and as the broader scientific and policy community develops frameworks for this space. We commit to updating this document when our thinking evolves — and to explaining clearly what changed and why. The version date and a full changelog are published on our GitHub repository.

Write to usRead the vision
VERSION · 2026.01 · LAST REVIEWED Q1 2026