Peer-Reviewed Journals Ran Papers With AI-Written Passages Nobody Disclosed

What happened

In March 2024, media organizations began reporting on a specific and growing problem inside peer-reviewed scientific literature: published papers containing text that had clearly been generated by a language model, with no disclosure from the authors that any AI tool had been used. The papers appeared across journals published by major academic houses, including Elsevier, and the evidence was often sitting in plain view, embedded in the prose itself.

The giveaway phrases were not subtle. Sentences like "as of my last knowledge update" and similar artifacts of a model hedging against outdated training data appeared in published introductions and methodology sections, language that no human researcher writes. These phrases do not indicate sophisticated AI-generated content that narrowly passed close scrutiny; they indicate text copied from a model's output with little or no review, then submitted and accepted through peer review without anyone noticing, or caring enough to flag it.

The failure runs in two directions. Authors used AI tools to generate passages they submitted under their own names, without disclosure, which misrepresents how the research was produced. But peer reviewers at high-profile journals cleared papers containing text that would have failed a basic quality check. Both failures compound each other: the ease of generating plausible-sounding academic prose made the first temptation larger, and the gap in reviewer scrutiny made the second one invisible until journalists found it.

The problem was not isolated. A 2023 Nature survey found that roughly 30 percent of the 1,600 scientists polled admitted using AI tools to help write manuscripts. That figure predates the March 2024 reports and suggests widespread undisclosed AI involvement in scientific writing, alongside a surge in retractions attributed to bogus or plagiarized studies. Related incidents included publishers withdrawing more than 120 gibberish AI-generated papers and a peer-reviewed journal publishing content that had no plausible human origin.

Scientific publishing depends on the premise that a paper's contents were produced, reviewed, and verified by the people whose names are on it. The March 2024 disclosures exposed how little of that premise can actually be checked after the fact. There is no mechanism at most journals to produce a provable record of what a system contributed to a submission, who reviewed that contribution, and whether it was disclosed before acceptance. Without that record, the integrity of peer review rests on trust that a fraction of the field has already decided to breach.

Reported impact

Affected parties

Not publicly disclosed

Harm type

Not publicly disclosed

Scale

Not publicly disclosed

Financial impact

Not publicly disclosed

Regulatory action

Not publicly disclosed

Classification

Organization

Not publicly disclosed

AI system

Not publicly disclosed

Industry

Not publicly disclosed

Country

Not publicly disclosed

Provider

Not publicly disclosed

Incident type

Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

• No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.