The '100% Human-Written' Medical Research Company That Was 100% AI

A company called EditMed, which marketed itself as providing "100% human-written, never AI" peer reviews for medical research, has been caught using AI to generate its content. The discovery, reported by 404 Media, exposes a growing problem in the academic publishing world: companies exploiting the AI backlash to sell "authenticity" while secretly using the very tools they claim to reject.

The Setup

EditMed positioned itself as the antidote to AI-generated content in medical research. Their pitch was simple: while other companies were flooding academic journals with AI-generated reviews and papers, EditMed offered "100% human-written, never AI" peer review services. For researchers and institutions worried about AI contamination in the scientific record, this was an appealing promise.

The company advertised that its reviewers were real humans who would carefully read and assess medical research papers. They charged premium prices for this "authentic" service. The marketing was effective — in an era where AI-generated content is increasingly difficult to distinguish from human-written work, a guarantee of human authorship has real value.

The Discovery

The investigation revealed that EditMed was using AI models to generate its "human-written" peer reviews. The content they produced — which they sold as the work of qualified medical professionals carefully reviewing research — was in fact generated by AI language models.

This is particularly damning for several reasons:

1. Medical research integrity: Peer review is the gatekeeping mechanism for scientific publication. If peer reviews are AI-generated, the entire quality assurance process is compromised.

2. False advertising: The company explicitly marketed itself as providing human-only content while doing the exact opposite.

3. Exploiting fear: They monetized legitimate concerns about AI in academic publishing by selling a false solution.

Why This Matters Beyond One Company

The EditMed case is a symptom of a larger problem. As the backlash against AI-generated content grows, so does the market for "human-only" and "AI-free" services. But without verifiable proof, these claims are just marketing — and the incentives to cheat are enormous.

This mirrors the "organic" food labeling problem. When consumers are willing to pay a premium for "organic" produce, some suppliers will label conventional produce as organic to capture that premium. The same dynamic is now playing out with "human-written" content.

Several platforms have emerged claiming to detect AI-generated text — GPTZero, Originality.AI, and others. But these tools have significant false positive and false negative rates. They cannot reliably distinguish AI-generated text from human-written text, especially when the AI output has been lightly edited.

The Bigger Picture for Academic Publishing

The academic publishing industry is already under strain from:

• Paper mills: Operations that produce fake research papers for sale
• AI-generated papers: Studies showing AI-written papers being accepted at conferences
• Peer review fraud: Reviewers colluding with authors to guarantee positive reviews
• Salami slicing: One study split into many small papers to inflate publication counts

The EditMed case adds a new dimension: fake human peer reviews. If the peer review process itself can be faked — either by AI or by fraudulent human reviewers — then the entire scientific publication pipeline is compromised.

What Can Be Done?

Several approaches could help:

1. Transparent AI policies: Journals should explicitly state whether AI tools are permitted in peer review and under what conditions, rather than banning AI outright (which creates the incentive to hide its use).

2. Reviewer verification: Journals could require more detailed reviewer profiles, including ORCID IDs and institutional affiliations that can be verified.

3. Open peer review: Making peer reviews publicly available alongside published papers would allow the community to scrutinize review quality.

4. Reproducible workflows: Using tools like EditMed's own AI detection (ironically) to flag suspicious patterns in reviews — while acknowledging these tools' limitations.

The Irony

The deepest irony of the EditMed case is that the company was caught by the same AI detection tools that its business model was built on avoiding. The technology used to identify AI-generated content is the same technology that EditMed claimed to never use — but did.

This suggests that the "AI-free" marketing trend may be a bubble. Companies that promise human-only content without verifiable proof are operating on borrowed time. Eventually, the gap between their marketing claims and their actual practices will be exposed — whether by investigative journalism, AI detection tools, or whistleblowers.

For now, the lesson is clear: in a world where AI can generate convincing content on any topic, the premium on genuine human work is real — but so is the temptation to fake it. Trust but verify, especially when someone is selling you "100% human."