Transparency

AI Editorial Policy

Updated: April 2026

Coherva uses artificial intelligence deliberately and extensively. This page explains precisely what AI does here, what it does not do, and the standards that govern every claim we publish. We believe transparency about AI use is not optional — it is the foundation of trust with a scientifically literate readership.

Our position is the inverse of most editorial AI policies. We do not use AI reluctantly as a productivity tool while aspiring to human-only authorship. We use AI as the synthesis engine for a publication whose core purpose is synthesis — connecting peer-reviewed neuroscience to 5,000 years of contemplative tradition in a form that is actionable, cited, and honest about what is known versus what is proposed.

What AI Does at Coherva

Four distinct functions, each with defined accountability

01 — Daily Resonance
AI synthesizes science and tradition daily
Each morning transmission is AI-authored from a curated research finding and its closest parallel in the primary wisdom texts. The scientific claim is sourced to a named, peer-reviewed study. The tradition parallel is drawn from a named primary text — not paraphrase, not interpretation, but the actual passage. A human editorial review is applied before publish.
02 — Research Archive
AI authors the synthesis essays from sourced research
Research articles in the Coherva archive are AI-authored synthesis pieces. Each article begins with a real peer-reviewed finding, maps it to its tradition parallels, and closes with the full citation chain. The AI writes the synthesis. The citations are the accountability. No article is published without named sources that readers can verify independently.
03 — Coherva Guide
Live AI in the dashboard for science and tradition questions
The Coherva Guide in the member dashboard is a live AI responding to questions about neuroscience, tradition, and practice. It identifies itself as AI. It draws on the same research base as the archive. It does not make health claims and does not replace professional guidance — this is stated explicitly at every entry point.
04 — Practice Generation
AI generates personalised practice sequences on request
Members can request a custom practice track for any theme — our AI generates a sequenced protocol grounded in the same neuroscience and tradition framework. Generated tracks carry the same PROVEN / THEORETICAL / EMERGING labelling as the core curriculum. The underlying practices are drawn from the documented archive, not invented.

What AI Does Not Do

The human editorial layer

AI is responsible for
  • Synthesizing science findings with tradition parallels
  • Authoring the connective prose of research articles
  • Generating practice sequences from the documented archive
  • Answering member questions in the Coherva Guide
  • Drafting daily transmission copy from sourced material
AI does not determine
  • Which peer-reviewed studies are cited — human editorial selection
  • PROVEN / THEORETICAL / EMERGING classification — human review
  • Health claims — none are made; content is educational only
  • Primary source selection from wisdom traditions — human curation
  • Whether a finding is ready to publish — human gate

The Classification System

Every claim is labelled. None are hidden.

PROVEN
Peer-reviewed and replicated. The finding is supported by published research in indexed journals, confirmed across multiple studies or by systematic review. A named citation is always provided. Example: cardiac coherence at 0.1 Hz produces measurable RMSSD increase — HeartMath Institute, McCraty et al. (2015).
THEORETICAL
Proposed mechanism, evidence-informed. The connection is coherent and draws on established science, but has not been directly confirmed by clinical trial. The underlying components are real; the specific link is proposed. Example: Hebbian consolidation applied to contemplative practice as a mechanism for lasting state change.
EMERGING
Active research, promising early evidence. Findings are preliminary — pilot studies, preprints, or early-phase trials. The direction is credible; the evidence base is still forming. We publish emerging research because excluding it would misrepresent the frontier of the field.

On synthetic citations: Coherva does not generate citations. Every source referenced in our archive is a real, named, verifiable publication. If a source cannot be independently confirmed by a reader, it is removed. AI hallucination of citations is a known failure mode of language models — our editorial process treats all AI-generated source claims as unverified until confirmed against the actual publication.

Our Commitments

Label everything. Every AI-generated synthesis carries its sources. The Coherva Guide in the dashboard identifies itself as AI at every point of interaction. No content arrives without a disclosed origin.

No health claims. Coherva content is for informational and educational purposes. The science we cover is real. The practices we document have documented effects. None of this constitutes medical advice, diagnosis, or treatment. This disclaimer is not legal boilerplate — it reflects an honest boundary between what peer-reviewed research demonstrates and what it licenses a practitioner to claim.

Human gate on publication. No article enters the public archive without editorial review. The AI produces a draft. A human reviews the citations, the classification, and the tradition parallel before the article is published. Volume does not override this step.

This policy will change. AI capabilities change rapidly. Our use of AI will evolve. When it does, this page will reflect it. The last updated date at the top of this page is current.