Research Gold, a medical research ghostwriting company that said it delivers "100% human-written" work and does not use AI, was exposed in a 404 Media investigation that found AI across its customer-facing operation, from phone support to team pages and quote emails.
404 Media found problems with both customer service and staff identities
According to the report, 404 Media reporter Emanuel Maiberg called the company and reached a voice identifying itself as an AI assistant named "Sarah." When he pressed on whether he was speaking to a real person, the response was, "Yep, I’m a real person," followed by renewed claims that the service was handled entirely by professional humans. The conversation then turned back to pricing.
The company’s "The Team" page raised more questions. The reported founder, Dr. Elena Vasquez, and eight other listed team members could not be verified, and their profile pictures appeared to be AI-generated.
Real researcher says her name and image were used without consent
404 Media said another section of the site listed methodologists who were not all fabricated. Some were real researchers, but their headshots had been taken directly from LinkedIn.
Jenny Berrio, one of the people listed on the website, said she only learned about it after being contacted by 404 Media. She said: "I do not work for Research Gold. I never consented to being listed as one of their methodologists, and I have no relationship with this company. They used my name, photo, and bio without my permission, and I am documenting this and will send a formal takedown request."
The report noted that a methodologist is responsible for checking whether a study’s process is rigorous. In this case, even the people presented as responsible for quality control were either invented or apparently misrepresented.
A test inquiry produced a full quote for $1,900
To test how the service worked, Maiberg submitted a fake request through the company’s form on the topic of "the impact of blogs on children aged 0 to 5," without supplying supporting material.
He quickly received what the report described as an AI-like email. It said children aged 0 to 5 are not "readers" in the usual sense and would need an operational definition or reviewers could object. It then asked whether the intended subject was how parenting blogs affect caregivers, or how digital content affects children themselves.
After Maiberg chose the latter, another reply followed quickly. The topic was broken into a PICO framework covering population, exposure, comparison group, and outcome measures. Soon after, the company sent a $1,900 quote covering proposal drafting, database searches, screening, extraction, risk assessment, and final formatting, along with a payment link.
No named response before publication
Maiberg later sent requests for comment. The response, again in an AI-style format, said the matter would be forwarded to a colleague who could speak on the record. Before publication, no one did.
BlockTempo first published the Chinese-language article summarizing 404 Media’s findings. The central issue in the report was not only whether Research Gold used AI, but whether the company’s claimed human team and represented service structure actually existed as advertised.