A firm selling ‘100% human-written, never AI’ systematic reviews had no humans in it

Research Gold, a company selling manuscript drafting, systematic reviews and meta-analyses to medical researchers, advertises its work as "100% human-written, never AI" and lists PhD methodologists on staff. 404 Media reported on 11 August that the staff do not exist.

What the reporting found

Reporter Emanuel Maiberg found that the eight people listed on the company's About page, including a "Founder & Lead Methodologist" credited with twelve years in evidence synthesis and a "Scoping Review Specialist," returned no online footprint matching their described careers, and that their profile photographs appeared AI-generated.

A separate section of the site listed a second group of methodologists whose photographs were real. Those were real people, taken from LinkedIn without their knowledge. One of them, evidence synthesis scientist Jenny Berrio, told 404 Media:

I do not work for Research Gold, and I never agreed to be listed as one of their methodologists. I have no relationship with this company.

At least one of the lifted photographs still carried the LinkedIn "#opentowork" overlay. That page was removed shortly after Berrio was contacted.

Attempts to reach a human failed in every channel. A phone agent introducing itself as Sarah insisted it was a real person and that the company was "all human expertise, all the way through." Email and chat replies also appeared to be AI-generated. The company did not provide a comment before publication.

The part researchers should sit with

Maiberg requested a quote for a systematic review of "the impact of blogging on ages 0-5," attaching no materials. The reply reframed the question into a PICO structure, noted that a 0 to 5 population is not a reading audience in the usual sense, offered two possible operationalisations, and on his choosing one, priced the full review at $1,900. That covered, in the service's own description, a registration-ready protocol, the search, dual screening, extraction, risk-of-bias appraisal, synthesis and a formatted manuscript. The site invokes PRISMA 2020 and the Cochrane Handbook.

Sebastian Rowan, a PhD candidate in civil and environmental engineering at the University of New Hampshire, who first flagged the company to 404 Media, described the alternative in concrete terms: he read over 250 articles start to finish for his own meta-analysis, and can cite specific references for every conclusion in his paper.

Analysis

The methodological reframing in that email is the detail worth noticing, because it is competent. It is the kind of pushback a good methodologist gives. It identified a real defect in the question and proposed two defensible ways to fix it. What it never did was decline.

A methodologist's value is not only in structuring a question. It is in being a person who will tell you your question cannot be answered, who is accountable for the search strategy, and whose name is attached to the appraisal. A PRISMA flow diagram records decisions somebody made. When nobody made them, the diagram is a picture of a process that did not occur, and it is indistinguishable from one that did.

There is a second-order problem here for the disclosure debate. Journals and funders are converging on rules that require authors to declare AI involvement. Those rules assume the author knows. Someone who commissions methodological support in good faith, and receives output from a firm that markets itself specifically on being human, cannot make an accurate declaration. The disclosure claim was the product.

Sources

• 404 Media, Emanuel Maiberg, Company Offering '100% Human-Written, Never AI' Medical Research Is Entirely AI, 11 August 2026
• Retraction Watch, Weekend Reads, 15 August 2026