For most of the history of journals, readers saw only the end product. The reviewers' objections, the authors' responses, the data behind the figures, the plan the researchers had before they collected any data, and the money that changed hands to publish the article all stayed out of view. That was not usually a matter of concealment. Print had no room for any of it, and the conventions of confidential review assumed that the process should stay private.
Over the past decade that default has been eroding, unevenly but steadily. Several separate movements — open peer review, data sharing, preregistration, disclosure of competing interests and price transparency — have each pushed more of the process into the published record. They have different champions and different motives: funders want accountability for public money, methodologists want protection against selective reporting, librarians want to know what they are paying for, and publishers want to show that their review processes are real.
This trend brief looks at where each strand stands, what is driving it, and where its limits lie. It is distinct from the question of how publishing infrastructure records corrections and identities, covered in JCF Media's explainer on designing trust into scholarly publishing.
Peer review comes out from behind the curtain
The most visible change is the publication of peer review reports. Nature offered authors the option of publishing reviewers' reports and author responses from 2020. From 16 June 2025 it made transparent peer review standard: every new primary research submission that is accepted is now published with its referee reports and the authors' replies. Reviewers stay anonymous unless they choose to be named. Nature Communications had moved earlier, offering the option from 2016 and applying it to all direct submissions from 2022.
The distinction matters. Transparent peer review publishes the content of review while usually preserving reviewer anonymity; open-identity review names reviewers. Many journals, including those at EMBO Press and PLOS, have offered one or both for years, often on an opt-in basis.
eLife went further. Announced in October 2022 and introduced in 2023, its model publishes every reviewed manuscript as a Reviewed Preprint, accompanied by public reviews and an eLife assessment summarising the significance of the work and the strength of its evidence. The editorial accept-or-reject decision after review is removed; authors decide when a version becomes the version of record.
Data availability becomes a condition, not a courtesy
Data availability statements are now common across many journals, but the stronger pressure comes from funders. The NIH Data Management and Sharing Policy, which took effect on 25 January 2023, requires a data management and sharing plan for NIH-funded research that generates scientific data, regardless of funding level. Individual NIH institutes may add their own requirements.
In August 2022 the White House Office of Science and Technology Policy issued the memorandum on Ensuring Free, Immediate, and Equitable Access to Federally Funded Research — often called the Nelson memo — directing US research funders to remove embargoes on publications and to make supporting data available. Its future is now uncertain. In February 2026 STM reported that the current Administration was reviewing potential options regarding the memo, and that congressional appropriators had asked OSTP to report on the status of efforts to repeal it. Publishers and institutions should track agency-level policies, which remain the binding instruments, rather than assume the memo's timetable still applies.
For journals, the practical shift is from asking authors whether data are available to checking that the statement points somewhere real, ideally a repository with a persistent identifier.
Preregistration and Registered Reports
Transparency can also run backwards in time. Preregistration — publicly recording hypotheses, methods and analysis plans before data collection — makes it possible to distinguish confirmatory analyses from exploratory ones. Registered Reports build this into the publication process itself. Authors submit their introduction and methods for review before collecting data; if the protocol is sound, the journal grants in-principle acceptance, committing to publish regardless of the results provided the approved plan is followed. According to the Center for Open Science, more than 300 journals now offer the format in some form.
The appeal is obvious in fields troubled by publication bias and underpowered studies. The limitation is equally clear: the format suits hypothesis-testing research and fits poorly with exploratory, descriptive or theory-building work.
Conflicts of interest and contribution
Disclosure of competing interests is long established, particularly in clinical research, but transparency here increasingly means structure rather than prose. Standardised disclosure forms and machine-readable contributor roles — the CRediT taxonomy became ANSI/NISO standard Z39.104-2022 — make it easier to see who did what and who might benefit. The weakness is that disclosure depends on honesty; journals rarely have the means to verify what authors leave out.
Price and service transparency
The newest strand concerns money. Under Plan S, cOAlition S requires that open access fees be commensurate with the services delivered, and it endorses two frameworks for publishers to report prices and services: Information Power's Plan S Price Transparency Framework, which breaks prices into eight service categories, and the FAIR Open Access Alliance's Publication Services and Fees framework. Data are collected through cOAlition S's Journal Comparison Service. Participation in the service is voluntary, though non-participation is shown in the cOAlition S Journal Checker Tool; OASPA has published practical guidance for publishers supplying the data.
Uptake has been uneven, and some publishers have argued that detailed price breakdowns are costly to produce or commercially sensitive. JCF Media's analysis of what an APC actually pays for explores the economics in more detail.
What this means in practice
- For editors, transparent review changes how reports are written. Reports that will be published are read by a wider audience than the authors and editor, and journals need clear policies on redaction and on what happens to reports for rejected papers.
- For publishers, every new transparency element is also a production and metadata task. Review reports need DOIs or stable links; data statements need validation; price frameworks need finance teams to allocate costs.
- For researchers, transparency adds work at submission and creates new forms of credit, such as citable review reports.
- For librarians and funders, the growing body of published review, data and price information offers a basis for assessing journals on their practices rather than their brand.
The limits of transparency
Transparency is a means, not a guarantee. Publishing a weak review does not make it strong, and a data availability statement that says data are "available on request" often proves hollow. There is also a real risk of overload: article pages carrying review histories, data links, disclosures and version notes can bury the findings readers came for. And transparency can create pressures of its own — for example, junior reviewers may be reluctant to criticise senior authors under open-identity models, which is one reason most journals publish reports anonymously.
Conclusion
The direction of travel is clear even where the pace is not. Published review reports, funder data mandates, Registered Reports, structured disclosures and price frameworks have each moved from experiment to established option, and in some cases to requirement. The policy environment, particularly in the United States, is less settled than it appeared in 2022. For publishers, the sensible response is to treat transparency as a set of capabilities to build deliberately — with clear policies, good metadata and readable presentation — rather than as a series of mandates to satisfy one at a time.
