The paradox of peer review: protecting science or policing thought?

Peer review is now so deeply embedded in the structures of science and scholarship that it is rarely questioned. It is widely regarded not just as a mechanism for assessing manuscripts but even as a defining feature of scientific knowledge itself. In academic life, scholarly success is often equated with numbers of peer‐reviewed publications, on which funding, career advancement and institutional recognition depend.

The underlying purpose of peer review appears clear and straightforward: it aims to bring together members of a community of inquiry who share relevant expertise and are prepared to engage with one another's work for mutual benefit in a spirit of critical reflection. Ideally, this process enhances the quality, rigour and reliability of scholarly production, supports the development of ideas and contributes to trust in the scientific record. In this sense, peer review is not merely a technical procedure but a form of collective action grounded in shared intellectual and ethical commitments.

These aspirations are compelling. They include the maintenance of epistemic integrity, the fostering of scholarly capacity, the preservation of a plurality of perspectives and the protection of public trust in knowledge. Around these aims an elaborate system of expectations has developed, supported by guidelines, protocols and institutional norms. Reviewers are expected to assess methodological soundness, identify potential misconduct, avoid conflicts of interest and maintain confidentiality. Journals and editors are tasked with coordinating and overseeing this process.

Yet despite its centrality, the effectiveness of peer review is less certain than is often assumed. Empirical studies suggest that, although it can improve aspects of reporting and presentation, the evidence that it consistently enhances methodological quality, detects error or guarantees the epistemic validity of published work remains limited and mixed. Peer review is widely treated as a safeguard of scientific integrity, but its actual capacity to fulfil this role is difficult to demonstrate and rarely scrutinised and, therefore, remains in doubt. ¹ , ²

At the same time, the conditions under which peer review operates have in recent years changed profoundly in response to transformations in the economic, institutional and technological environments of scholarly communication. ³ As a result, the reviewing system itself must now confront significant challenges.

The first set of these is economic. Scholarly publishing, once largely overseen by learned societies and academic communities, is now highly concentrated within a small number of large commercial enterprises. These organisations control a substantial proportion of journal output and often report profit margins that exceed those of many other industries. ⁴ Under subscription‐based models, this has contributed to escalating costs and restricted access, particularly for institutions and researchers in less well‐resourced settings.

The shift towards open access has, if anything, exacerbated the inequities. While it has expanded the availability of published work to readers, the widespread adoption of ‘article processing charges’ has transferred the financial burden from institutions and libraries to authors and their funders. As a result, the capacity to publish increasingly depends on authors' access to resources, thereby naturally favouring the voices of scientists and academics in wealthier settings rather than reducing well‐recognised inequalities in the global distribution of knowledge. In this sense at least, open access, once presented as a democratising reform, has actually become the opposite. ⁵ , ⁶

Within this framework, the costs associated with peer review itself – particularly its administration and coordination – are incorporated into publication charges, even though the labour of reviewers themselves is typically unpaid. This creates a further tension: a system that relies on voluntary scholarly contribution is embedded within an increasingly commercialised model of knowledge production.

A second set of challenges arises from changes in the institutional environment of universities and research organisations. Reviewing was once recognised as an integral component of academic life, supported – if only informally – by institutional cultures that valued collegial contributions. In many settings, however, increasing managerial demands, higher teaching loads, and greater precarity of employment have reduced both the capacity and the incentive for academics to participate in reviewing. What was once understood as a shared professional responsibility is now frequently experienced as an additional burden. ⁷

Editors commonly report that the recruitment of appropriate reviewers has become the rate‐limiting step in the publication process. The difficulty lies not only in numbers but in the availability of reviewers with the requisite expertise, judgement and willingness to engage constructively with the work of others. In response, there are emerging efforts to formalise and, in some cases, to remunerate aspects of the reviewing process, including platforms designed to match reviewers to manuscripts and provide recognition and compensation. While these developments remain in an early stage, they raise important questions about the potential commodification of peer review and the possible perverse incentives that may thereby be generated. ⁸ , ⁹

A third set of challenges arises from the rapid development of artificial intelligence. Large language models now make it possible to generate text that resembles scholarly writing with relatively little input. This has implications for authorship, originality and accountability. At the same time, similar technologies can be used to generate or assist peer reviews, potentially increasing efficiency but also risking the substitution of procedural compliance for genuine critical engagement. While the long‐term impact of these technologies remains uncertain, they introduce additional ambiguities about what constitutes meaningful acts of scholarly judgement. ¹⁰

Taken together, these developments contribute to a fourth and even more difficult set of concerns: the transformation of the culture of reviewing itself. The mere existence of a review process does not ensure that it is conducted in a manner that is expert, disinterested and constructive. Nor does it guarantee that the broader aims of peer review – supporting truth, fostering innovation and enabling critical dialogue – are realised in practice. For these aims to be achieved, reviewers must be true peers: individuals who share relevant knowledge and expertise and who are prepared to engage with openness, humility and intellectual generosity. Reviewing, in its ideal form, is a dialogical practice. It involves not only evaluation but conversation – an engagement with uncertainty, ambiguity and possibility. In this sense, it is inherently ethical: it depends on mutual recognition, respect and a shared commitment to the advancement of knowledge and to deeper understanding. ¹¹

Many participants in the contemporary system, however, report a different experience. Reviews may be brief, directive and occasionally dismissive; requests for revision can reflect strong preferences for particular frameworks or interpretations; and anonymity, while supposedly offering protection, can reduce accountability. Empirical studies have documented the presence of unprofessional or biased commentary in many reviews, with potential effects on authors, particularly those from underrepresented or less well‐resourced backgrounds. These phenomena are not universal, but they are sufficiently common to warrant concern. ¹² , ¹³ , ¹⁴

At the same time, the increasing reliance on guidelines, checklists and formalised procedures, while intended to standardise and improve review, may have only accelerated these regrettable tendencies further. By emphasising compliance with predefined criteria, such instruments risk narrowing the scope of evaluation and displacing the reflective, interpretive and dialogical elements that are, or should be, central to the reviewing process. What was conceived as a collaborative engagement oriented towards understanding may, in many contexts, be reduced to a process of verification and control.

None of these problems, of course, is entirely new. From its earliest forms, peer review has been open to bias, error and misuse. What is distinctive about the present predicament is the convergence of multiple pressures – economic, institutional and technological – that together reshape both the function and the meaning of reviewing. Under these conditions, the system has become increasingly strained.

Whether these trends can be reversed or redirected is uncertain. Several possibilities may be envisaged. Academic institutions could recognise and support reviewing as a core scholarly activity. Mechanisms could be developed to mitigate financial barriers to publication, particularly for those without access to substantial resources. Greater transparency in the reviewing process, including forms of open or dialogical review, might enhance accountability and foster more constructive engagement. Editors and editorial boards could play more active roles in guiding reviewers and shaping the tone and substance of reviews, including by moderating inappropriate or irrelevant commentary.

As unlikely as these changes may seem, however, even more is needed. The ethical purpose of peer review also needs to be rearticulated and refreshed. This would involve a recognition that reviewing is not simply a technical filter but, more fundamentally, a practice embedded in a broader moral and social order. It is concerned not only with the evaluation of individual manuscripts but also with the cultivation of a community of inquiry in which ideas can be tested, challenged and developed.

This, then, is the current paradox. Peer review was established to protect the integrity and vitality of scientific discourse but under contemporary conditions risks becoming a mechanism that constrains innovation, reinforces orthodoxy and amplifies the inequalities of the system within which it operates. We therefore now need to ask not what further guidelines are needed to regulate peer review but what form peer review should take: Can it be restored as a practice of scholarly responsibility and dialogue, grounded in shared ethical commitments, or will it continue to erode into a procedural mechanism shaped by market forces, technological systems and institutional imperatives which reproduces and exacerbates global inequities? The answer to this question will determine not only the fate of peer review but also the future of scientific and scholarly communication itself.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.