The first thing to understand about the World Congress on Research Integrity 2026 is that the programme is not really about what researchers do wrong. It is about what universities, funders, and the journals those universities lean on for prestige refuse to change when nobody forces them. The 9th edition of the congress opens in Vienna on 24 May 2026 and runs through 27 May, hosted by the Austrian Agency for Research Integrity (OeAWI) with the World Conferences on Research Integrity Foundation (WCRIF). The accepted abstract list reads less like a victory lap and more like an audit of the field's unfinished business: paper mills, machine-generated evidence, research assessment, and the gap between signed declarations and local hiring practice.
The congress series is not a minor gathering. Since the first meeting in Lisbon in 2007, it has produced documents that funders and universities quote in policy: the Singapore Statement on Research Integrity in 2010 and the Cape Town Statement on fairness in research collaborations in 2022. Vienna is the next test of whether those documents can become infrastructure rather than decoration.
The abstract list is the nearest thing the field has to a risk register
A first read of the accepted abstracts shows five recurring fights. Organisers have clustered sessions around what the field knows is broken, not around what looks good in a brochure. Among the strongest clusters:
- AI-generated text and image manipulation in manuscript submission, including tools that flag doctored western blots and the countermeasures that follow.
- Paper mills and the limits of editorial screening, from detection guidance to publisher-level mass retractions.
- The reform of researcher assessment and the slow fate of the Hong Kong Principles, which were drafted after the 2019 congress but remain unevenly applied.
- National and funder integrity frameworks in the Global South, where reporting capacity and legal protection still determine what gets investigated.
- Reproducibility, open data, and the difference between depositing a dataset and making it usable by another lab.
The same themes were present in Athens in 2024. Their return is not a failure of imagination. It is a signal that the problems are structural, which means they do not disappear just because a consensus forms in a plenary hall.
AI and paper mills: the abstracts nobody should call solved
If there is a single throughline in the 2026 abstracts, it is that artificial intelligence has made fabrication cheaper and detection more expensive. A clutch of accepted talks examines image duplication in oncology and cell biology papers, where small alterations can produce a retraction years after publication. The paper-mill problem itself is not new, but the abstracts focus on a newer fact: generative AI can now produce plausible data tables, clinical summaries, and referee reports, which shifts the burden of proof from editors to institutions that often lack forensic capacity.
COPE, the Committee on Publication Ethics, has published specific guidance for spotting and handling paper mills, and several sessions build directly on that framework. The Retraction Watch database remains the most visible public record of how many published papers have been pulled. Conference abstracts are not enforcement, though. Naming a manipulation technique in Vienna does not give a research integrity officer the authority or budget to open an investigation back home.
The strongest sessions do not treat detection as a technical race. They ask why a promotion committee still counts a retracted paper as a valid line on a CV and why a publisher takes months to act after a clear paper-mill pattern is reported. For people who track editor-level decisions, COPE's paper-mill detection guidance became a reference point long before this congress; the Vienna abstracts are, in effect, a set of field notes on where that guidance has failed.
Photo by Krzysztof Hepner on Unsplash
Here's the catch
The programme is good at naming failures. It is less good at naming who will be accountable for fixing them. A congress can recommend that universities stop using journal-level metrics in hiring; it cannot make a dean remove an impact factor from a tenure committee checklist. A congress can expose a paper mill network; it cannot make a university pursue a misconduct finding against a senior author if the legal office refuses. The gap between the best abstract and a changed departmental rubric is where research integrity actually lives.
History supports the scepticism. The Hong Kong Principles were adopted as a model for responsible researcher assessment in 2019 and promptly collided with short hiring timelines and long CVs. The Singapore Statement was signed and then ignored by institutions that never built investigation procedures. Vienna will produce new commitments. The question is whether anyone in the room has the authority to bind a promotion panel, a publisher, or a government funder in the weeks after the closing session.
Hype level: high, as it should be for the largest integrity meeting of its kind. Reality level: institutions can nod in Vienna and change nothing in their next recruitment cycle.
What the assessment reform abstracts ask of employers
Several accepted sessions take aim at the instrument panel that still shapes academic careers: the journal impact factor, the h-index, and the publisher prestige ladder. The abstracts are not asking for a new metric. They are asking for slower, more structured reading of actual research outputs. That means panels would need to read a few papers properly rather than counting forty titles, a demand that collides with the eleven-minutes-per-file reality many hiring committees describe.
The tension is visible in the programme's treatment of journal indexation. Some abstracts argue that indexation has been mistaken for a quality guarantee, a point AcademicJobs has covered in its reporting on journal indexation as a shortcut. The practical question after Vienna is whether universities will redesign their academic calendars and hiring workflows to make qualitative review possible. Most employers already have the data to do this; they rarely have the time, the training, or an administrator willing to shield the extra reading time.
For early-career researchers, the stakes are direct. A programme that says promotion should reward transparency, open data, and rigorous null results means little if postdoctoral contracts still run on two-year cycles and require a specific paper count. The abstracts are full of better rubrics. The bottleneck is not design; it is implementation.
The Global South abstracts refuse to be a side event
One of the more consequential shifts in the accepted abstracts is the number of sessions that treat research integrity as a question of resourcing, not just individual misconduct. Researchers from lower-resourced systems describe the same problems as their better-funded colleagues, but with fewer trained integrity officers, weaker legal protections for whistleblowers, and publisher pricing that determines which literatures they can even read. The Cape Town Statement's emphasis on equity in research collaborations gets more direct engagement in Vienna than many past congresses managed.
Several abstracts connect paper mills to the economics of authorship in systems where a publication can determine a salary band or visa eligibility. The point is not to excuse fabrication. It is to explain why an enforcement-only response will keep missing the parts of the system that produce demand for fake papers. A congress that finally treats the Global South as a set of research governance systems, rather than as a compliance problem, is a different kind of meeting. For those unfamiliar with the series, the World Conferences on Research Integrity Foundation publishes past statements and congress proceedings.
Photo by Markus Spiske on Unsplash
What to watch when the closing session ends
The real output of Vienna will not be a communiqué. It will be the set of questions that follow: whether any university changes a promotion rubric in the next academic year, whether a publisher tightens its paper-mill triage in response to a specific abstract, and whether funders add research integrity capacity to grants rather than another declaration to their websites. The programme already tells us what the field knows. The post-congress behaviour will tell us what the field is willing to act on.
That is exactly the standard Brian Nosek has applied to the reform of scientific incentives for more than a decade at the Center for Open Science: don't judge a reform by how good it sounds, judge it by whether the reward system changes enough for a researcher to behave differently without being punished. Vienna will give the research integrity community plenty to say. The hard part is what happens when the audience goes home and the next hiring packet hits the desk.