Research Gold has received questions about how our company operates: whether our research deliverables are written by people, whether artificial intelligence (AI) was involved in our client communication, and whether the team profiles previously shown on our website were genuine. Recent reporting raised these questions publicly, and our clients deserve direct answers rather than marketing language.
This page is that answer. It states plainly what has always been human at Research Gold, what was automated, what we got wrong, and what has changed.
The research deliverables are, and have always been, human work
Every client deliverable produced by Research Gold, including systematic reviews, meta-analyses, statistical analyses, and manuscript documents, is researched and written by human experts.
A serious evidence synthesis project involves framing the research question, defining eligibility criteria, building and running search strategies, screening records, extracting data, assessing risk of bias, selecting and running the statistical synthesis, and interpreting the results. Those decisions are made by people, and the documents delivered to clients are composed by people.
We stand behind this, and behind the work itself.
What was automated, and what we got wrong
Parts of our front office were automated. Automated assistants helped handle initial website chat, some phone enquiries, and some early email exchanges, including initial project scoping.
Two things about that were wrong, and we say so without qualification:
- The assistants did not clearly identify themselves as automated. Visitors who asked whether they were speaking with a person were not given a straight answer. That should never have happened, and it is not how we want to treat the researchers who contact us.
- Our marketing used the phrase "100% human-written, never AI" without making clear that it described the research deliverables, not every system used to run the business. Applied to the deliverables, the claim was accurate. Applied to the whole company, it was not.
We have discontinued the automated assistants. Today, when you contact Research Gold, you are corresponding with a person, whether by chat, email, or phone.
The team page previously shown on our website included profiles that should never have been published. Some did not correspond to a verifiable person, and some used material belonging to professionals who had not agreed to appear on our site. That was wrong, and we do not defend it.
We removed the team page. Profiles will return only when the identity of each person has been verified and each person has directly confirmed their consent to appear. No exceptions.
The experts who perform client work are real. The failure was in how our website represented people, and our responsibility is to fix that representation, not to defend it.
The research community is actively working out how AI involvement should be described and declared. We published our own guide on declaring AI use in systematic reviews, covering the thresholds journals apply, and we are now holding ourselves to the same precision. Our commitments:
- Human deliverables. Client research is researched, written, and checked by people. We do not deliver machine-generated research.
- Human communication. Client conversations are with people.
- Clear disclosure. Where software assists a workflow, we describe exactly what it does instead of reducing it to a slogan. Our screening platform is a good example: it offers optional AI suggestions during record screening, those suggestions are never applied automatically, and a human reviewer makes every call. That design is documented on the tool itself.
- Verified identities. Anyone represented as part of Research Gold will be a verified, consenting, real person.
We are not asking anyone to accept these statements on trust alone.
Our free research tool library is public and open to anyone: PRISMA flow diagram builders, search strategy tools, risk of bias templates, statistical calculators, and many more. Our systematic review screening platform can be examined directly, including how its optional AI suggestions are built around human final decisions. And how we run client projects is described step by step, using the same methodology our published guides teach.
Technology can support a research workflow. It does not replace the people responsible for the work, and it will not at Research Gold.
If you are a client, past or present, and want to know how your project was handled, get in touch and we will tell you directly.