Methodology & research integrity
Methodology, sources, and independence.
The rules for everything Enpath publishes: how evidence is treated, how it is cited, how it is corrected, and how independence is protected — including our own.
Methodology
The Enpath Standard.
The Standard is a simple test: what an AI answer must show before it's ready for a real decision. If it can't answer these, it isn't ready.
-
Why was this recommended?
The reasoning is visible, not just the conclusion.
-
What evidence supports it?
Sources are cited and can be opened.
-
What did it assume?
Assumptions are stated, not hidden.
-
What is still uncertain?
The unknowns are named, with how much they matter.
-
Who is accountable?
A named person, identified before publication.
-
How can it be challenged?
There is a real way to question or correct it.
-
How can it get better?
What would change the answer, and how it improves.
The bright line
If those seven questions can’t be answered, the system isn’t ready for a decision — no matter how convincing it sounds.
Sources
How we handle evidence.
- 01
Every claim carries its source
Named and linked wherever possible. Fact, analysis, and assumption are kept apart.
- 02
Uncertainty is stated
Where evidence is thin, we say so in the text. "We don't yet know" is a finding.
- 03
Primary sources
We cite the study, dataset, or institution of record — not a headline about it.
- 04
Estimates are labelled
Modelled or extrapolated figures are marked, with their vintage and assumptions.
Corrections
We correct in the open.
Substantive corrections — a changed figure, a withdrawn claim — are logged on the piece itself, with the date and what changed. Minor fixes are made without a note.
Found an error? Tell us. Reports are read and answered by a person.
Published: the correction, its date, and what changed
Independence & funding
No funder directs a conclusion.
Enpath is funded by its advisory work for governments and companies, its training, and its independent audits. No funder, client, or partner directs our conclusions, and any funding relevant to a piece is disclosed on it.
Where a piece touches a domain in which Enpath has a commercial interest, that interest is disclosed. We publish no paid content or sponsored research.
July 9, 2026
AI audits. Earlier versions of this site said Enpath would not audit third-party AI. We now offer independent audits under three rules without exception: public methodology, non-negotiable findings, and no audited organization may fund our research.
Published: relevant funding, on every publication
Our own use of AI
Held to our own Standard.
- AI tools assist our research and drafting.
- Every publication is reviewed and fact-checked against its sources by a person, who answers for it.
- No piece is published on a model’s authority: a human verifies the sources first.
- Our own work is held to the same seven questions of the Standard.
Published: who reviewed and verified each piece
See the Standard applied.
To quote or republish our work, with attribution and a link to the original, write to us. Cite as: Enpath, "[title]," enpath.co, [year].
Last reviewed: July 2026