This page explains, in one place, how content on GlobalFXHub actually gets made: how we research and source it, where AI is and isn’t involved, how we handle corrections, how we stay independent of the brokers we cover, how scores are calculated, and what we do when sources disagree. See How we test for the scoring formula itself — this page is about process.
Broker entries are compiled from public regulatory registers (CySEC, the FCA, the Seychelles FSA, and others), each broker’s own disclosures, and third-party broker-comparison and review sites, cross-referenced against each other wherever possible. Where sources conflict or a fact couldn’t be independently confirmed, we say so directly in the text rather than presenting one source’s claim as settled fact — you’ll see phrasing like “not independently confirmed,” “sources disagree,” or “a single, unverified account” throughout our broker pages for exactly this reason.
We use AI assistance in both research and translation: compiling and cross-referencing publicly available broker information, and translating our content into other languages (reviewed before publishing). AI assistance speeds up research and keeps our disclosed methodology applied consistently across every broker, but it does not invent facts — where a figure can’t be confirmed from an actual source, our content says so instead of guessing, and a human reviews and can correct anything that’s wrong. We do not publish fabricated people, credentials, or experience as bylines; see the “Authors” section below.
Content not otherwise credited is attributed to the GlobalFXHub Research Team rather than to an invented individual persona. If we name a specific person as an author in the future, that will be a real person who actually did the work, with a real, verifiable background — never an invented name, photo, or biography.
We correct mistakes when we find them — a wrong licence number, an outdated regulatory status, a stale fee figure — and keep researching rather than leaving a page wrong because it’s already published. Material corrections (ones that would change what a reader took away, as opposed to a typo fix) are logged at our corrections log with what changed and why. See also our automated change log, which separately tracks when our own published pricing, platform, or licence data actually changes.
GlobalFXHub does not currently have referral or advertising relationships with any broker listed on this site. If that changes, we’ll disclose it clearly — see Advertiser Disclosure — and it will not change how a score is calculated. Our scoring formula is fixed and disclosed in advance; it is never adjusted for a specific broker.
Every broker is scored against the same disclosed, absolute rubric — see How we test for the full breakdown of all nine categories and their weights. Scoring is based on public regulatory data and disclosures, not hands-on account testing. A red-flag finding (a withdrawal complaint, a conduct allegation) only affects a numerical score when it’s corroborated by an attributable source — a regulator action, a named tracker’s own adjudication, a settlement or fine — never from an anonymous or single-source complaint alone.
Where two credible sources give different figures for the same fact, we say so explicitly in the text (“sources disagree: one cites X, another cites Y”) rather than silently picking one. Where we can’t independently confirm a figure at all, the relevant scoring category either scores a neutral midpoint or is excluded from that broker’s average entirely, rather than guessed at — see How we test for exactly how that works.
Questions about any of this? Contact us via the contact page.