Member states have almost settled on a call to make immediate open access the default, with no author fees. But some say the Council needs to do more to prevent AI-generated papers threatening the integrity of the scientific record
Research ministers are nearing the finish line in drafting their position on changes needed in academic publishing, with the latest leak draft revealing the near-final text.
The upcoming paper, to be adopted by ministers in late May, will call on policymakers and publishers to make immediate and unrestricted open access, “the default mode in publishing, with no fees for authors.”
While there is a warm reception for this, there is concern that the Council is missing the opportunity to crack down on AI-generated scientific articles, given rising evidence that the AI chatbots like ChatGPT could undermine the integrity of academic publishing.
The move to make research free at the point of publication builds on years of work in pushing open access policy in the EU, which has already seen many science funding agencies and organisations steering researchers they fund to publish their work in paywall-free journals, allowing results to be widely disseminated.
The Swedish presidency of the EU council first presented the draft conclusions in February and member states have been negotiating the details of the text since then.
With the negotiations almost done, the text hasn’t change too much since the original Swedish presidency draft. Overall, it’s balanced, although some believe it doesn’t go far enough to guard against AI generated papers and paper mills that fabricate research results.
The newest version makes clear the member states’ ambition to crack down on predatory publishing, which sees dodgy platforms exploiting the academic need to publish research. These publishers will publish anything for a fee or trick researchers into publishing in poor quality journals, often misleading younger scholars from developing countries.
The conclusions, the member states hope, take a clear stance against such practices and mandate the European Commission to start addressing the issue.
Universities are happy EU governments have responded to their plight. “I am very positively surprised by the Council, because the last amendments have taken a strong position, really highlighting the problems we experience as universities,” says Julien Chicot, senior policy officer at the Guild of European Research-Intensive Universities.
Mattias Björnmalm, secretary general of the university association CESAER, says he welcomes the direction of the conclusions, but there’s room for further action, in particular when it comes to dealing with synthetic AI text and moving towards fully endorsing secondary rights to republish publicly funded research after its first publication in an open access repository or elsewhere.
“On the one hand, I warmly welcome the attention and advancement here. On the other hand, we’d really like to move towards the boldest conclusions that we can have,” says Björnmalm.
The big ambition in the conclusions is the call to tackle the cost of open access publishing. Giving free access to scientific knowledge is getting more and more expensive, as rather than journal subscriptions, publishers charge fees for handling manuscripts, called article processing charges (APCs). It’s a hot issue for science at the moment – last week, the journal Nature reported that more than 40 editors from leading Elsevier-owned neuroscience journals had recently resigned in protest at unsustainable publishing fees.
Publishing costs money, but many believe the current APC rates can far exceed the actual cost of publication. The member states want to stop this, as it deepens inequalities among researchers and institutions, with many being unable to afford APCs. Universities welcome the ambition. “That is a very strong statement for us. That is what we are experiencing,” says Chicot.
Björnmalm adds it’s important that governments also recognise a full transition to open science won’t be free and resources will need to be dedicated to it.
It’s also about policy change. In a move welcome by universities, the conclusions recognise the introduction of secondary publication rights in a number of member states as an achievement. These enable publicly funded research to be republished in open access repositories even if they’ve been published in traditional journals before.
Björnmalm says this is an important addition to the conclusions. “It’s about giving new rights and empowering researchers,” he notes.
Chicot said member states should work together to ensure secondary publishing rules are harmonised across Europe. Otherwise, fragmented legislation “could hinder the development of these kinds of rights,” he warns.
The next step will be finding ways to translate ambition into action. Both member states and the research community hope the discussions can continue in the European Research Area Forum, which brings together the Commission, EU governments and research groups to discuss policy.
“Hopefully we can also continue the discussion in the ERA Forum and ERA action sub-groups, and together with the Commission and member states try finding a solution,” says Chicot.
The last technical discussion on the draft are expected to take place in the Council this Thursday, and there some still hoping to make a change or two to the text before it lands on the ministers’ tables for signing.
A group of countries hope to make a last-minute addition recognising multilingualism as an enabler for open science. While in life sciences and science, technology, engineering, and mathematics (STEM) the lingua franca is English, much of social science and humanities research is published in national languages. The current conclusions fail to recognise the importance of disseminating research outputs to everyone, in a variety of languages.
Björnmalm says a big missed opportunity – if no changes are made on Thursday – is taking a stance on the rise of AI-generated scientific articles. Since the AI chatbot ChatGPT launched in November, anecdotal evidence of its impact on academic publishing has been accumulating, signalling that it could become a major challenge soon.
“A worst-case scenario here is a Library of Babel situation where the scholarly publication system is overwhelmed by this type of synthetic texts and images, leading to a complete breakdown of the integrity of the scientific record,” warns Björnmalm.
To avoid disaster, Europe should lead by example and start solving the issue early, Björnmalm notes. First, by acknowledging that resources will be needed to safeguard the integrity of the scientific record and then by taking action at EU level.
“It’s a question of whether we want to shape developments or lag behind and fix things,” says Björnmalm. “Either we show leadership now or we wait and risk it becoming a much bigger problem, and then having to make a much bigger effort later on.”
Some member states reportedly campaigned to add a reference to AI-generated content in the conclusions, yet it didn’t make it to the final text. Discussions are likely at a later stage, as governments acknowledge AI will radically change the status quo in academic publishing.