The peer review process has been the cornerstone of science dissemination for centuries. ^{1 , 2} It exists in different forms, ^{3 , 4} all based on the elusive concepts of referees’ impartiality ⁵ and competency. ⁶ The peer review process is known to be slow, expensive, inconsistent, biased, ^{3 , 7 – 11} and with presumably limited impact on the quality of research publications. ^{12 , 13} These limitations have led to some calls to completely abandon the whole process because “ prepublication peer review is an enormous sink of scientists’ time, effort, and resources”. ¹⁴ Despite the limitations, one can note that the research community has still strong faith in it ^{7 , 15} ; i.e. “ despite the limitations of peer review process, we need it. It is all we have, and it is hard to imagine how we could get along without it”. ¹⁶ How to reform the peer review process has always been an extremely complex endeavor. ^{17 , 18}

Many calls have been made to reform the peer review process, typically along four schools of thought according to Waltman and colleagues (2023). Here, this paper subscribes to the Efficiency & Incentives school that focuses on streamlining the peer review process and incentivizing participation in it. ¹⁹ This is a completely pragmatic choice because without an efficient peer review system, there is no point in my opinion in discussing its quality, reproducibility, transparency, or inclusion. Those qualities, promoted by other schools of thought, are extremely important but they are contingent on the suitability of the existing system to cope with the current research productivity levels of millions of scientific articles. ²⁰ Specifically, over 2.5 million papers are published each year in English alone, ²¹ creating an incredible burden on editors and reviewers. The problem is obvious to all stakeholders: the exponential increase in the number of manuscripts ^{22 , 23} largely surpasses the current availability of qualified referees. Put another way, the current peer review process is not well equipped to serve modern publication dynamics that are expected to accelerate even more in this era of generative AI.

Adequate and sustainable alternative solutions are needed. As nature abhors a vacuum, if the current peer review model is not reformed, frustrated authors will have to rely on other means to get their studies out, in particular when desk rejection is reaching high rates. ^{24 , 25} This might for instance lead to a surge in alternative models based on the publication of unreviewed preprints, or the creation of local journals or repositories owned and managed by the authors’ universities and research institutions. ²⁶ However, the biggest risk is the undesirable proliferation of predatory journals offering a “pay-to-publish” model sometimes without peer review. ^{27 – 29} I believe that models based on pre-publication peer-review are still useful as they do serve the interest of science relatively better than other models. ¹⁸ In that context, I discuss here the idea of promoting the role of professional reviewers, like professional editors, who will be monetarily compensated for their role in the peer-review process.

Below, I describe a concrete situation based on my own experience as editor-in-chief of a scientific journal. The journal I am editing, as might be the case for other journals, is experiencing a (welcome) growth in the number of submissions that are putting too much pressure on an already overstretched peer review process. ²⁵ Consequently, this situation has resulted in (i) an unethical increase in desk rejection rates, (ii) a low percentage of invited researchers willing to review papers, (iii) an increase in the number of papers rejected after being sent for peer-review due to the difficulty to secure two or more referee reports, (iv) referees not submitting their reports after initially agreeing to do so, (v) referees submitting scant reports, listing superficial, unhelpful, or even irrelevant comments and suggestions, and (vi) an increase in instances of fraud where for example false contact information of potential referees with non-institutional emails was suggested or when the peer review process was manipulated. ^{30 – 32} All these issues are directly or indirectly related to the pressure to ensure a fair and rigorous peer review process when the main players (reviewers) are no longer motivated to give away their time and expertise for free.

In this context, a wide range of solutions and innovations were proposed in the past (see analytical review in ³³ ) to address the non-availability of volunteer peer reviewers. ^{18 , 33 – 39} Table 1 provides a succinct summary of some proposed innovations in the current literature. This includes, for instance, a more efficient involvement of editors in selecting prospective reviewers, ⁴⁰ improving diversity, ⁴¹ training reviewers, and opening up the peer review process. ⁷ Others have argued that scientific peer reviewers are rejecting invitations not necessarily because of lack of time but presumably because of lack of motivation. ⁴² Therefore, rewarding reviewers can help increase their (extrinsic) motivation, like offering certificates, discounts on publisher’s products, and publicly acknowledging the best and the most productive reviewers. ^{42 , 43}

However, these suggestions might not solve existing problems for five main reasons: (i) the number of submissions is projected to increase at much faster rates in this era of generative AI, (ii) the roles and duties of the academics include administration, writing grants, writing papers, writing patents, commercialization ventures, leaderships and directorships, editorships, teaching, supervision, committee memberships, consultancy, and outreach, leaving no time for reviewing papers (iii) the number of journals and publications in languages other than English is growing fast, promoted by current initiatives to improve inclusion in science dissemination, which makes it challenging to ensure a rigorous multilingual peer review process, (iv) researchers avoid peer-reviewing because they are increasingly unsatisfied ‎with for-profit publishers that reap exorbitant ‎profits while contributing negligibly towards academia and research, and (v) perhaps most importantly, researchers are likely to become less willing to spend time on reviewing AI-enhanced quick-to-write papers (e.g. when the time taken to review a paper becomes longer than the time taken to draft that paper with AI), leading to a growing reliance on generative AI by referees for peer review. I believe the latter is likely to be increasingly present in the debate about reforming the peer review process, which is why monetary incentives might be one potential means to alleviate the burden on the peer review process to some extent.

Publishing research articles is still an expensive business, even in this digital era. ^{44 – 46} The global annual cost of peer review is estimated at around $1.5 billion, ⁴⁷ corresponding to a net annual contribution of over 100 million hours by reviewers. ⁴⁸ In the current science dissemination business, which is evaluated at billions of dollars, scientific peer reviewers still provide volunteer labour. Journals should thus develop methods to compensate reviewers for their time while maintaining the integrity of the peer-review process. ^{47 , 49} The system needs fair and sustainable mechanisms to pay referees. This would allow researchers to earn money by becoming full-time or part-time professional reviewers. Like the remunerated role played by professional language editors, researchers who enjoy evaluating research papers can be invited to dedicate more time to the peer review process for a fee. ⁵⁰ For instance, a professional reviewer can be tasked with reviewing a given number of papers per week (e.g. 2 or 3 papers/week), which would help speed up the peer review process, in particular for journals with large submission volumes.

We already have models where referees are paid for reviewing academic work, albeit at a smaller scale than article journals, as grant reviewers, PhD examiners, or academic promotion examiners. Reviewing articles is no exception and could thus be considered a remunerated task in academia. There is room for the emergence of two different but complementary professional reviewers. Reviewers can still be academicians, i.e. tenured staff and academics, working in universities and research institutions, where they can be offered the opportunity to supplement their salaries with additional honorariums for their active involvement in the peer review process. Another alternative concerns the possibility of outsourcing the peer review process by allowing the creation of independent providers. These providers, managed by for-profit companies, ⁵¹ can follow the same work model as companies that employ professional language editors in research dissemination. The job of a professional reviewer can be regulated by adopting new ethical and professional standards to safeguard the peer review process from any possible unethical practices that might occur when money is involved in the process ( Figure 1). Furthermore, peer-reviewing is also undertaken by non-tenured staff and academics with relevant ‎qualifications and expertise (i.e. postdocs, senior technical staff, research assistants, laboratory engineers and managers, ‎ and PhD-qualified persons who have quit academia). However, they generally cannot claim academic promotions for their voluntary participation in ‎the peer-review process. Thus, an inclusive peer review process must acknowledge their contributions by financially rewarding ‎them.

The proponent of this idea argues that peer review should be considered as a business transaction, ⁵² where a modest remuneration per paper of around $200, ⁵² or $450 for for-profit publishers, ⁵³ can be offered to reviewers. This can motivate reviewers to be more actively involved in the peer review process and to offer fast and thorough review reports. ^{54 , 55} This can also be an incentive for retired scientists ^{52 , 56} to participate in peer review to gain an extra income that expands their retirement plans. Paying referees can increase the pool of available reviewers, including for instance researchers who cannot afford to work for free. ⁵⁶ Many factors may interplay to determine the level of remuneration, ^{48 , 57 , 58} including reviewing for not-for-profit versus for-profit journals, scientific domains (e.g. research in medicine is more funded and expensive than research in art and humanities), the number of reviewers required per manuscript (e.g. two or more referees), the type of articles (e.g. reviewing a research article requires more hours than a commentary), and their length and complexity (e.g. reviewing a long review article or a full research article with complex data is more demanding than an opinion piece). Publishers and editors should estimate the level of remuneration according to the average hours typically spent by referees for each article type ⁴⁸ and the average yearly number of manuscripts reviewed by each reviewer. ^{39 , 58} There is also the question of the type of financial compensation to referees: e.g. The ResearchHub Journal has recently started paying referees in cryptocurrency, ⁵⁹ PeerJ journal offers tokens that can be redeemed against publication fees, and Biology Open journal started offering real money at the level of $300 per manuscript for freelance reviewers. ⁶⁰ To guarantee fair and ethical payment schemes, active organizations and agencies should establish general guidelines about the payment of reviewers, including the Committee on Publication Ethics, the Declaration on Research Assessment, the Confederation of Open Access Repositories, and the European Association of Science Editors. Below, I discuss some options to sustain the payment of professional referees.

Reviewers as consultants: Researchers are usually invited to budget for the support of a consultant or a technical expert when applying for grants. For instance, the consultant can provide expert opinion about an aspect of the research project that might be beyond the expertise of the researchers (e.g. a legal advisor, an industrial partner, a clinical professional, a computer system expert, a programmer, … etc). In the same way, a reviewer can be seen as a consultant whose role is to support the researchers in improving their papers in terms of readability, quality, and methodological soundness. Publications are typically included as deliverables in research grants, and therefore, the contribution of scientific peer reviewers to improve these deliverables can be budgeted as consultancy fees. Funding bodies may categorize this type of expense as publication fees. However, regardless of how paying reviewers is included in the budget, too much financial burden will be put on funding bodies as they already cover the excessive article-processing charges. Similar to some alternatives being sought for article-processing charges, ⁶¹ funding bodies should put in place guidelines to make the payment process of professional reviewers as fair and transparent as possible, particularly when for-profit agencies are managing the peer review process. For instance, many funding bodies already require that published work be made freely available, and thus similar requirements can also be made about the type of peer review (free or paid). If there is strong evidence that paying reviewers improves the quality of publications, then funding bodies can recommend researchers to publish in journals that pay referees. This ensures that funded research is being evaluated and scrutinized by the most rigorous review process that scholarly communication can offer (and afford).

Levying a submission fee: As is the case in some journals, a submission fee can be collected for each submitted article, which can then be used to pay professional reviewers. For example, around 45 of the economics, finance, and accounting journals published by Elsevier charge a submission fee of around $50-$100. Likewise, some biology journals also levy submission fees. Their rationale is to ensure that only within-scope and high-quality papers are received. Submission fees also serve other purposes such as sponsoring conferences and workshops, the payment of language editing services, prizes/awards such as the annual best paper award, and even payments to top reviewers. To make the payment of professional reviewers sustainable, it might be suggested to levy a submission fee of $200 per submitted paper. For example, for a journal with a submission volume of 1,000 papers per year, and a desk-rejection rate of 50%, the collected submission fees (around $100,000) can help partially cover the costs of professional reviewers. To make this option attractive for authors, the submission fee should be deduced from the article-processing charges or the open access fee once the peer-reviewed paper is accepted. Consequently, this might mean that publishers should consider reducing their article processing charges when levying a submission fee. However, this might sound unsavory for for-profit publishers and hence the community needs to create an ecosystem that would make publishers active contributors to peer review, as discussed below. Likewise, having a submission fee that is “lost” upon rejection may deter authors from submitting to journals with up-front costs. This is why it would be more practical to integrate the submission fee into the overall publication charges applied by journals. Lastly, levying a submission fee might not be feasible for diamond open-access journals that do not charge authors and readers, and thus, they may still have to rely on voluntary referees.

Paying referees in the context of economic disparity: One of the consequences of paying reviewers might be a better diversity in peer review. This is because paying referees might be very appealing for researchers from low- or middle-income countries. However, this monetary incentive can inadvertently yield over-involvement in the peer review process because a fee of $200 or more is sometimes comparable to a full salary in some low-income countries. Put another way, researchers from low-income countries can favour paid peer review over other academic duties, which might not be a tenable situation for their institutions. One way to avoid this race for more reviews, and hence more money, is to link the number of invitations for referees from low-income countries to the number of authors or readers from those low-income countries. For instance, if a journal has a readership of 20% from low-income countries, the number of invited referees from low-income countries can be capped at twice this percentage. This measure would also prevent for-profit publishers from taking advantage from the existing diversity by increasing reliance on referees from low-income countries and providing them with inadequate (low) compensation. Furthermore, all paid review tasks should be declared on the publisher’s website so that editors are well-informed about the level of involvement of different referees. This would help identify unusual review dynamics that need adjustments to ensure a fair and robust peer review.

How about the publishers? The research community has voiced clear dissatisfaction about the ‘parasitic’ behaviour of some for-profit publishers in research dissemination. They are seen as the biggest winners of the current system despite not funding research and not paying the original content creators (the authors) and evaluators (the reviewers). ⁶² Recent examples of editors quitting for-profit journals provide an interesting illustration to appreciate the depth of that dissatisfaction (e.g. Refs. 63, 64). I believe that, if for-profit publishers do not engage with the research community more transparently and constructively, they will soon be made irrelevant to the whole process. Publishers must realize that they cannot continue operating along the same model. For instance, they can inject some of their profits back into the peer review process to support the remuneration of reviewers. However, it is unlikely that publishers would be willing to contribute financially to the peer review process unless they have control over the process through the creation of their professional review agencies. Still, I don’t think the research community can again trust for-profit publishers with the peer review process, which is understandable given the previous experience of the community with the open access model that has unfortunately been made so expensive and unaffordable by some for-profit publishers.

In this context, I feel that scholarly communication might foreseeably follow a radical model by completely bypassing for-profit publishers, i.e. a model that does not need the ‘middleman’. This implies that researchers and their universities will own the whole process. ⁶⁵ In this model, universities and research institutions might choose to manage the whole publication process, for instance in the form of posting papers on open repositories, ⁶⁶ while adopting an open post-publication peer review model. Universities can work as consortia to share the cost of publication through common or shared repositories, which can be further facilitated by the adoption of generative AI in the editorial process. ^{67 , 68} The other implication here is to allow researchers (or their institutions) to keep the copyright of their work instead of conceding it to publishers as in the current model.

The influence of journal indexers: In addition to publishers, journal indexing bodies can play an important role in shaping the future of peer review (e.g. Web of Science, Scopus). Those journal indexers have the power to influence how journals are conducting peer review. Specifically, journal indexers assess the impact of various peer review systems on the quality of published papers. This assessment can lead to delisting journals that do not meet standards of scholarly communication, such as editorial integrity, peer review quality, and ethical publishing practices. The case of the journal eLife, removed by Web of Science from its Science Citation Index Expanded, offers an interesting example. The journal eLife, in addition to its embedded consultation model among peer reviewers, implemented a new (bold) publishing model called “Publish, Review, Curate”, in which manuscripts are first published as preprints and subsequent peer review reports and editorial assessments are published alongside the original preprints and revised articles. ⁶⁹ However, Web of Science concluded that the new model conflicted with its quality assurance standards, leading to the journal not receiving an impact factor rating from Web of Science in 2025. ⁷⁰ This also concerns many journals removed from the Scopus database for various reasons, including questionable peer-review processes.

Indexers are crucial in maintaining quality standards, such as delisting journals that manipulate the system or fail to offer rigorous peer reviews, as is often the case with many predatory journals. However, it is equally important for indexers to proceed cautiously in this regard, ensuring they do not hinder innovation in the peer review process by allowing sufficient time for new initiatives to prove their value, including the idea of paying reviewers. Overall, indexers should maintain an honest and pragmatic dialogue with all stakeholders to ensure that their objectives align with those of the scientific community, beyond financial or reputational considerations.

Arguments against paying scientific peer reviewers: The opponents to paying reviewers mention the risks of proliferation of unethical practices (i.e. the system being corrupted by money) and unsustainable costs ^{53 , 56 , 71} as it can lead to increases in subscription fees and article processing charges. ⁷² Moreover, adding a monetary component to the peer review might lead to unethical practices to maximize the number of completed reviews, such as non-disclosure of conflict of interest, the tendency to review papers even outside one’s domain of expertise, and a fast turnaround time with no thorough evaluation of the reviewed papers. Therefore, journals should put some safeguards to ensure a high quality and rigorous peer review process, for instance, through the evaluation/rating by the editors of the quality of completed reviews before approving payment to the referees. However, such quality criteria are unfortunately still missing, given also the fact that that the expertise of reviewers and their personal style to provide feedback differs greatly. One potential alternative is to rely on structured peer reviews when implementing any payment schemes. Overall, perhaps the recent experiment at the journal Open Biology might provide a concrete example that peer review quality with paid reviewers can be guaranteed through a rigours editorial oversight and review. ⁷³

It is also likely that a flat fee (e.g. $200 per manuscript) for all review rounds might create conflicting interests between reviewers and editors. Reviewers may be inclined to reject papers with extensive revisions, while editors prefer multiple revisions over rejections and resubmissions to avoid additional payments. Similarly, reviewers might not provide high quality feedback for additional review rounds when compensation is offered per manuscript. Journals may also invite fewer reviewers to reduce costs. Again, all these concerns underscore the need for robust safeguards to mitigate their negative impact on the overall quality of peer review.

Likewise, paying reviewers might yield a proliferation of for-profit peer review agencies with dubious ethical practices, like paper mills, that can generate fabricated review reports. ⁷² Some studies have argued that monetary rewards tend to decrease the quality of the peer review process ⁷⁴ as well as the intrinsic motivation of researchers to contribute to the peer review process. ⁷⁵ Paying reviewers might be only sustainable for journals with a reasonable submission volume. ⁷⁶ For multidisciplinary journals with broad topics and large submission volumes, the peer review process can be enormously costly and thus diverse funding sources have to be sought. There is also the risk that paying referees will soon become the norm, making researchers no longer interested in reviewing for free. It could also lead to a multi-tier peer review system depending on how much a journal chooses to pay its reviewers, hence engendering an unequal competition among journals for the service of reviewers. This could hurt not-for-profit publishers that cannot afford an expensive peer review process. In that context, not-for-profit publishers must (i) diversify incentives on top of paying reviewers, (ii) nurture the intrinsic motivation of their readers to participate in peer review and serve their scientific community, ⁷⁷ (iii) seek alternative means to be subsidized, ⁷⁸ and (iv) foster strategic partnerships with all stakeholders to make not-for-profit journals the preferred model for open-access publicly-funded research. Overall, although one cannot rule out the risk of corruption in peer review by money, I trust that academia has built-in mechanisms to handle financial transactions fairly and objectively, as is the case when researchers are paid to evaluate grant proposals and academic theses. ⁵⁶

Other schemes of financial compensation: It might be worth mentioning other alternative models. For example, one proposal argues for offering monetary compensation not for the reviewers themselves but for their institutions ⁷⁹ as they usually deal with the same publishers for purchasing subscriptions and other products. The rationale is that reviewers are already paid by their institutions, and hence, their involvement in the peer review process should benefit their institutions. However, this sometimes ignores the fact that researchers sometimes (often?) conduct their reviews outside work hours (e.g. evenings or weekends), and hence their contributions to peer review should not be assumed already covered by their salaries. Likewise, another recent proposal argues for monetary prizes not to individual referees but to their groups (departments, research units or labs). ⁷⁵ That collected money, for instance, can support ongoing research projects, support early career members or the work of members from underrepresented minorities. ⁷⁵

The above-mentioned solutions are assumed remedies to a not-fit-for-purpose peer review process. However, it is possible that the peer review process itself is adequate, and that the challenges encountered come from the environment within which peer review is conducted. Indeed, the culprit here is the current academic environment that promotes an unhealthy culture of “publish or perish”. To survive in this culture, researchers might fragment their results into multiple publications or even publish redundant papers based on similar data. ^{37 , 80} This culture of “publish or perish” is hurting academic research. ⁸¹ Therefore, instead of instigating calls to abandon or reform the peer review process, academia should invent and campaign for new models that depart from the “publish or perish” model. ⁸² As long as research quality is measured as a function of research productivity, there will be no form of peer review that can guarantee quality research while dealing with a deluge of papers that is now being inflated even more by the adoption of generative AI.

Is the ultimate goal of academia to produce prolific researchers who can publish several papers a week, ⁸³ sometimes with questionable practices? ⁸⁴ If, hypothetically, this goal is achieved by every researcher, who would review (and read) the published work? Evidence shows that many published papers remain uncited or end up having little impact on their respective fields, ⁸⁵ but one cannot ignore the resources those papers took (time and effort) from reviewers and editors. It is as if the whole purpose of many substandard papers is to inflate some research metrics. Can academia promote a new model that puts a cap on how many papers an author can submit per year? This is essentially to invite researchers to think sensibly about the quality of their research and maybe only publish the best of their research that is likely to have an impact on their respective fields. Other work can still be shared as unreviewed preprints. ⁸⁶ This framework ensures a manageable number of submissions, though it might entail a rethinking of current citation-based metrics of research impact and productivity.

We can’t afford to ignore existing flaws in the current peer review system because the status quo will only favour the emergence of unethical practices. Here, I support the idea, albeit not new, of creating professional peer reviewers. Paying reviewers must be made fairly and transparently. Future work needs to assess if paying reviewers has a positive impact on the quality and rigor of peer review. Perhaps the recent “Fast & Fair peer review” experiment, run by the journal Biology Open where reviewers are paid for their work, offers some encouraging results about the positive impact of paying reviewers on the quality of peer review. ⁶⁰ This experiment needs to be systematically tested in other journals and domains. Future work also needs to investigate alternative ways to ensure a sustainable and affordable remuneration of professional reviewers. The discussion should continue among all stakeholders to identify new solutions and innovations for efficient and affordable research dissemination in the modern era of generative AI. Empowering universities and research institutions to own the whole research dissemination system might soon be an inevitable scheme to consider. Finally, in the absence of effective and bold academic reforms, scholarly communication will continue relying on a costly and ineffective peer review system that will sooner or later be completely abandoned or replaced by something else.

This opinion article was conceived and written by MLS.