CASRAI Dictionary

Category: Policy & Funding News

Reporting and briefings on external policy, regulatory, and funder developments affecting the research community worldwide.

  • REF 2029 Portability: What Changes for Moving Staff

    REF 2029 portability ends automatic transfer of most research outputs when staff change institutions. Outputs are now “decoupled” from individual authors and instead require a demonstrable “substantive link” to the submitting institution, with one exception: long-form outputs such as monographs stay portable with the author for five years. The rule, confirmed by the four UK funding bodies in December 2025, replaces the dual-submission compromise used in REF 2021.

    Portability, in REF terms, is the rule set that determines which institution — the one a researcher has left, or the one they have joined — may submit a given research output for assessment. REF 2029 narrows that rule further than any previous exercise, and the change has become one of the most contested elements of the framework’s redesign.

    What REF 2029 Changes for Portability

    REF 2029’s initial decisions, published in 2023, “decoupled” research outputs from the individual researchers who produced them. Where REF 2021 attributed outputs to named staff on a submission list, REF 2029 requires only that a submitting institution demonstrate a substantive link to the output — evidence that it supported the underlying research.

    Research England’s Head of REF Policy has described this as a shift toward assessing “how well organisations are supporting research excellence” rather than tracking individual output counts. In practice, that means the REF volume measure and staff lists are no longer the sole determinants of which outputs an institution may submit.

    For standard outputs — journal articles, conference papers, most datasets and software — the substantive link is generally established by an eligible employment relationship between the institution and the author when the research was conducted or the output first made publicly available. Once a researcher leaves, that link does not automatically transfer to their new employer.

    This is a marked departure from REF 2021, which let both the origin and destination institution return an output where a researcher transferred, and from REF 2014, where only the destination institution could submit if the move happened before a single census date. The table below sets out how the rule has moved across three cycles.

    REF cycle Portability rule Long-form treatment
    REF 2014 Single census date; only the destination institution could submit if a researcher moved before it No separate treatment
    REF 2021 Compromise rule: both origin and destination institutions could return the same output on transfer No separate treatment
    REF 2029 Outputs decoupled from individuals; standard outputs require a substantive institutional link and are not portable Portable for five years if the new post began within that window

    The REF 2021 compromise itself had a contested history: Lord Stern’s 2016 independent review of the REF recommended full non-portability, but Research England judged that too significant a change to introduce that late in the REF 2021 cycle, and adopted dual submission instead.

    Why Long-Form Outputs Get a Five-Year Exception

    On 10 December 2025, following a three-month pause and intense sector pushback, REF 2029 published updated guidance restoring limited portability for “long-form and extended-process” outputs — monographs, edited collections and scholarly editions. These remain attached to the author for five years, provided the researcher’s employment at the new institution began within that window.

    According to Times Higher Education’s reporting on the announcement, institutions can deploy this exception where the academic was employed at some point during a two-year staff census window that opened in September 2025 — though commentators on that same report disputed how directly the census window bears on long-form eligibility, underlining how unsettled the operational detail still is.

    Reaction split along familiar lines. Jennifer Richards, chair of the English Association and professor of English literature at the University of Cambridge, called herself “delighted that an element of portability has been restored,” describing it as “a ‘win’ for all researchers in a sector that has become increasingly precarious at every stage of career.” Margot Finn, vice-president for higher education and research at the British Academy, called the concession “excellent and very welcome.” Rosa Freedman, professor of law at the University of Reading, was less convinced, arguing the “half in-half out” approach “doesn’t seem to have much logic” and that outputs should belong either wholly to the employer or wholly to the researcher.

    What the Sector Consultation Flagged as Unresolved Risk

    Sector consultation ahead of the December 2025 decision surfaced risks that the rule change does not fully resolve. Institutions can still retain long-form outputs credited to staff they have since made redundant — a concession the reversal did not touch. Several concerns recur across the consultation record:

    • Early-career and precariously employed researchers lose standard outputs as a bargaining tool when applying for new posts, since those outputs no longer travel with them.
    • Redundant staff may see their long-form outputs retained and submitted by an institution that has already ended their employment.
    • Co-authored outputs where authors move to different institutions raise unresolved questions about which institution’s substantive link takes precedence, particularly where authorship attribution is contested.
    • Discipline-specific impact: arts, humanities and social science bodies — including the English Association, the Institute of English Studies and University English — wrote jointly to flag that decoupling disproportionately affects fields where long-form outputs and precarious contracts are both common.

    A joint Wonkhe analysis by academics at the Royal Central School of Speech and Drama, the University of Oxford and the University of Edinburgh went further, warning that decoupling removes any auditable limit on how many outputs by one individual a unit can submit, which they argued undermines the diversity-of-contribution goal the rule change was partly designed to serve.

    Common Questions on REF 2029 Portability

    What period does REF 2029 cover?

    REF 2029 runs on an extended timetable after the exercise was renamed from REF 2028 to REF 2029 in a December 2023 decision. The criteria-setting phase runs to summer 2026, the assessment phase is planned for winter 2028 to autumn 2029, with results published in December 2029.

    What outputs are eligible for REF 2029?

    Standard outputs are eligible where the submitting institution demonstrates a substantive link — typically an eligible employment relationship with the author when the research was conducted or first made public. Long-form outputs, such as monographs, remain eligible at a new institution for five years after the researcher’s move.

    What is the weighting for REF 2029?

    REF 2029’s December 2025 guidance weights Contributions to Knowledge and Understanding at 55% of the overall quality profile, Engagement and Impact at 25%, and Strategy, People and Research Environment at 20%, shifting emphasis toward institutional research culture rather than individual output counts.

    Is it REF 2028 or REF 2029?

    It is REF 2029. The four UK funding bodies renamed the exercise from REF 2028 to REF 2029 in December 2023, extending the assessment timetable to give institutions more time to adjust to decoupling, open access and portability rule changes ahead of results in December 2029.

    What Institutions Should Do Next

    Research administrators handling staff transfers should treat the substantive-link test, not the researcher’s current employer, as the operative question for every output under consideration. That means auditing when each output’s underlying research was conducted or first made public, and mapping it against the employment record at that time — a task that sits squarely with research administration teams managing REF submission pipelines.

    For long-form outputs, institutions should document the five-year window and the justification for classifying a work as “long-form and extended-process” at the point of hire, not retrospectively. Because the rule change also reopens questions about how authorship attribution interacts with institutional submission rights, joint appointments and co-authored outputs deserve particular scrutiny before the assessment phase begins in 2027.

    The sector consultation record makes clear that portability remains an unresolved fairness question, not a settled technical rule. Institutions that build audit trails now — rather than waiting for further REF guidance — will be better placed when panels begin assessing submissions from 2028 onward.

  • BioRxiv Preprint Server: NIH, Wellcome and Gates Compared

    Funder preprint requirements diverge sharply in 2026: the Gates Foundation and the Howard Hughes Medical Institute (HHMI) now mandate deposit on a recognised server such as the preprint server bioRxiv or its sister site medRxiv, the NIH indexes eligible preprints in PubMed Central without treating them as compliance, and Wellcome requires them only in defined public-health-emergency scenarios. Research administrators tracking multi-funder portfolios need a single reference for which rule applies where.

    A preprint server is an open-access repository — such as bioRxiv for biology or medRxiv for health sciences — where researchers post a complete but not-yet-peer-reviewed manuscript for immediate public access. Both platforms are operated by openRxiv, a nonprofit formed in 2025 specifically to run bioRxiv and medRxiv independently of their founding host institution.

    What Is a Preprint Server Like bioRxiv?

    A preprint server is a repository for manuscripts that have not yet completed formal peer review. bioRxiv, co-founded by John Inglis and Richard Sever, launched in November 2013 as a life-sciences equivalent to arXiv. Its companion site, medRxiv, covers health and clinical research and applies additional pre-posting screening because of the sensitivity of medical findings.

    Submissions to both platforms undergo basic scrutiny — plagiarism screening, an appropriateness check, and a safeguarding review — but not peer review itself. Roughly two-thirds of bioRxiv preprints are later published in a peer-reviewed journal, and by early 2026 bioRxiv was recording around four million article downloads a month, according to a Nature analysis of the server’s first 13 years.

    Which Funders Require Preprinting in 2026?

    Funder policy on preprints splits into three tiers: outright mandates, conditional requirements, and pure encouragement. The table below summarises the position of four major research funders as of 2026.

    Funder Preprint requirement Effective date Compliance role
    Gates Foundation Mandatory — deposit before or at journal submission, CC BY licence 1 January 2025 Core requirement of the Open Access Policy
    HHMI Mandatory for HHMI investigators, scholars and Janelia scientists 1 January 2026 Preprint required before journal submission
    Wellcome Trust Required only for research with significant public health implications; encouraged otherwise Ongoing Accepted as a fallback open access route if a fully OA journal or Europe PMC deposit is unavailable
    NIH Not required Preprint Pilot ongoing Discoverability only — does not satisfy the NIH Public Access Policy

    Two funders — Gates and HHMI — now treat preprinting as a compulsory step in the research lifecycle. NIH and Wellcome instead fold preprints into a wider menu of open access routes, which is the detail most other coverage of this topic omits.

    How Does the NIH Preprint Pilot Treat bioRxiv Deposits?

    The NIH Preprint Pilot makes eligible NIH-funded preprints — including bioRxiv and medRxiv deposits — discoverable in PubMed Central and PubMed, tagged clearly as preprints rather than peer-reviewed literature. This is a discoverability mechanism, not a compliance mechanism.

    Posting a preprint does not fulfil the NIH Public Access Policy. Grantees must still deposit the final, accepted peer-reviewed manuscript in PMC. NIH does, however, permit researchers to cite preprints in grant applications and progress reports, which gives early findings some formal standing without changing the underlying compliance obligation.

    Does Wellcome Require or Just Encourage Preprints?

    Wellcome strongly encourages preprinting across its funded portfolio and requires it specifically where a disease outbreak or comparable public health emergency makes rapid sharing a priority. Outside those defined scenarios, preprinting is not compulsory.

    Wellcome’s primary open access compliance routes are publication in a fully open access journal or deposit of the author’s accepted manuscript in Europe PMC. A CC BY-licensed preprint is accepted as a valid compliance route only if neither of those primary routes is available — a fallback position, not a default requirement.

    What Changed When bioRxiv and medRxiv Became openRxiv?

    bioRxiv and medRxiv were hosted by Cold Spring Harbor Laboratory (CSHL) from launch until 11 March 2025, when ownership transferred to openRxiv, a newly formed nonprofit dedicated solely to running the two preprint servers. This is a provenance detail funder-policy roundups routinely miss, and it matters for research administrators: openRxiv, not CSHL, is now the governing body whose terms of use and licensing options apply to deposits made under Gates, HHMI, Wellcome and NIH-linked research.

    The Chan Zuckerberg Initiative has funded platform development on both sites since 2017, part of a broader pattern of philanthropic infrastructure investment that runs parallel to — and distinct from — the funder mandates covered above.

    Common Questions About bioRxiv and Preprint Servers

    Is bioRxiv a preprint server?

    Yes. bioRxiv is an open-access preprint server for the biological sciences, hosting complete but not-yet-peer-reviewed manuscripts. It does not conduct peer review itself, though submissions undergo basic screening and reviews from journals or platforms such as Review Commons may be posted alongside preprints.

    Is it free to publish on bioRxiv?

    Yes. Authors register without charge and there is no fee to deposit a manuscript on bioRxiv or medRxiv. This zero-cost deposit model is one reason funders increasingly treat preprinting as a low-friction first step toward full open access compliance.

    What is a preprint server?

    A preprint server is an online repository where researchers post manuscripts before or during formal peer review, giving the wider research community immediate access to findings. bioRxiv and medRxiv are the leading discipline-specific examples in the life and health sciences.

    What are the disadvantages of preprints?

    Preprints add an extra step to the publishing process, are not peer-reviewed at the point of posting, and can attract premature media coverage or public comment before findings are validated. Some journals and funders still weigh these risks against the benefit of faster dissemination.

    What This Means for Institutions and Researchers

    Research administrators managing grants across multiple funders now need to track preprint policy at the individual-funder level rather than assuming a single institutional rule applies. A grantee funded jointly by Gates and NIH, for example, must preprint to satisfy Gates while still separately depositing the accepted manuscript in PMC for NIH.

    • Confirm licence requirements before deposit — Gates and HHMI specify CC BY, which is not the default licence offered on every server.
    • Do not treat NIH Preprint Pilot indexing as equivalent to Public Access Policy compliance — the two are separate obligations.
    • Check Wellcome’s public-health-emergency criteria before assuming preprinting is optional on a given grant.
    • Record the openRxiv terms of use in grant files, since bioRxiv and medRxiv are no longer governed by Cold Spring Harbor Laboratory.

    The Direction of Travel for Funder Preprint Policy

    The trend across 2025 and 2026 runs firmly toward mandatory preprinting among the largest philanthropic funders, while NIH and Wellcome hold a more conditional position rooted in their existing open access frameworks. Institutions should expect more funders to follow the Gates and HHMI model as preprint infrastructure matures under openRxiv’s independent stewardship. Research administration teams that build funder-specific preprint checklists now, rather than applying a single blanket policy, will be better placed as more mandates convert from encouragement to requirement.

    For related definitions and terminology used across research administration and open access compliance, see the CASRAI Dictionary and the research administration resource hub.

  • bioRxiv License Update: What Changed for Authors and Reuse

    bioRxiv’s licence update, live via the platform’s Author Area since January 2026, lets authors request a change to a less-restrictive Creative Commons licence on a preprint already posted — without submitting a new version. The change can only move in one direction, towards more permissive reuse, and it exists chiefly to help authors bring older preprints into line with funder mandates that require CC BY.

    bioRxiv is a free preprint server for the life sciences, operated by the non-profit organisation openRxiv, which also runs the companion server medRxiv for health-sciences preprints.

    Contents

    What is the bioRxiv licence update, and why was it introduced?

    The bioRxiv licence update is a self-service feature that lets a preprint’s corresponding or submitting author switch its Creative Commons licence to a less restrictive option after posting, without triggering a full revision. openRxiv documented the mechanics in a step-by-step guide published on 7 January 2026, and followed up with a policy explainer on 20 May 2026 setting out the rationale.

    The trigger is compliance drift. Openrxiv’s own explainer states that a growing number of funders “require their grantees to apply specific licenses to their preprints, typically CC BY,” but that “many authors are unaware of this” and post under a more restrictive option by default. Before this update, the only remedy was submitting an entirely new version of the preprint and re-selecting a licence — a heavier process that also generates a fresh revision record. This is distinct from an earlier, smaller change in January 2025, when bioRxiv and medRxiv reordered their licence-selection menus to place CC BY at the top of the list; the 2026 update is the first mechanism that lets authors retroactively fix the licence on preprints they have already posted.

    How do authors request a licence change?

    The workflow runs entirely through the bioRxiv submission system’s Author Area and does not require re-uploading a manuscript. It applies only to the most recent version of a preprint, and only to preprints posted within the past two years.

    • Log into the Author Area from the bioRxiv submit page.
    • Locate the “Request License Update” box on the right-hand side of the page.
    • Select “Update license choice on previously posted papers.”
    • Choose the eligible preprint by its manuscript ID (only papers where the requester was corresponding or submitting author are listed).
    • Select a new, less restrictive licence and submit the request; a confirmation email follows.

    Two constraints apply strictly. First, the feature is unavailable if an incomplete revision is already in the submission system, or if a previous licence request is still pending. Second, a request can even be made after the preprint has been formally published in a journal, since the licence sits on the preprint record independently of the journal’s own copyright terms.

    Licence options compared: what actually changed

    bioRxiv preprints have long offered a choice of Creative Commons licences plus a “no licence” (all rights reserved) default, and a CC0 public-domain option for US federal employees such as NIH intramural researchers. What changed in 2026 is not the menu of options — it is that authors can now move an already-posted preprint from a more restrictive option to a less restrictive one after the fact.

    Licence Commercial reuse Attribution required Text-and-data mining / AI training Typical funder fit
    CC BY Permitted Yes Unrestricted, including commercial use HHMI, Gates Foundation, most cOAlition S funders
    CC BY-ND Permitted (no derivatives) Yes Mining permitted; no adapted/derivative outputs distributed Rarely funder-compliant
    CC BY-NC Not permitted Yes Restricted to non-commercial use Non-compliant with CC BY mandates
    CC BY-NC-ND Not permitted Yes Most restrictive; non-commercial, no derivatives Rarely funder-compliant
    CC0 Permitted (public domain) No Unrestricted US federal/NIH intramural authors only
    No licence selected Not permitted without separate permission N/A Reuse requires author permission Non-compliant with most funder mandates

    Because Creative Commons licences are irrevocable once attached to a public copy of a work, the update only runs in the permissive direction. An author can move from CC BY-NC to CC BY; the system rejects a request to move from CC BY to a more restrictive licence, since existing downloaded and archived copies would remain under the original, broader terms regardless.

    What this means for CC-BY reuse, text-and-data mining, and AI training

    bioRxiv’s baseline terms of use already permit text-and-data mining of posted content, which is the legal hook that has made preprint corpora attractive training data for machine-learning systems. The licence attached to an individual preprint then determines the scope of onward reuse beyond that baseline — and this is where the 2026 update has practical bite.

    Under CC BY, any party — including a commercial AI developer — may reproduce, adapt, and redistribute the work, provided the original authors are credited. Under CC BY-NC or CC BY-NC-ND, commercial reuse (which covers most AI model training conducted by for-profit developers) is not licensed, regardless of the platform-level text-mining consent. That gap is precisely what several funders have moved to close: the Howard Hughes Medical Institute’s preprint requirement, effective 1 January 2026, and the Bill & Melinda Gates Foundation’s preprint mandate, in force since 1 January 2025, both require grantee preprints to carry CC BY. The licence-update feature exists to let authors already out of step with those mandates fix a specific preprint without a full resubmission.

    For institutions and research-integrity offices, the practical implication is that a preprint’s licence — not merely its posting on an open server — is the operative variable for downstream reuse and AI-training permissions. Auditing grantee preprints for licence compliance, not just for the fact of preprint deposit, is now a distinct compliance step.

    Answer-first Q&A

    How do I update a bioRxiv?

    Authors can request a licence update from the Author Area of the bioRxiv submission system, using the “Request License Update” box, without submitting a full revision. The change applies only to preprints posted in the past two years and only to the most recent version, moving to a less restrictive licence.

    What are the licence options for bioRxiv?

    bioRxiv authors can choose CC BY, CC BY-ND, CC BY-NC, CC BY-NC-ND, or leave the preprint with no licence (all rights reserved). A CC0 public-domain option is also available specifically for US federal employees, such as NIH intramural researchers.

    Does bioRxiv count as published?

    No. A bioRxiv preprint is not peer reviewed and does not constitute formal journal publication; it is a publicly posted manuscript with its own DOI. Authors remain free to submit the same work to a journal afterward, and the preprint record persists independently of that later publication.

    Who maintains bioRxiv?

    bioRxiv is operated by openRxiv, a non-profit organisation dedicated to advancing science communication, which also runs the companion health-sciences server medRxiv. openRxiv is supported by institutions including Cold Spring Harbor Laboratory, the Chan Zuckerberg Initiative, and the Sergey Brin Family Foundation.

    Implications for institutions, funders, and authors

    Research-administration offices tracking open-access compliance should treat the licence update as a remediation tool, not a substitute for correct licence selection at submission. It closes a specific gap — preprints posted before an author understood their funder’s CC BY requirement — but it does not apply to preprints older than two years, to superseded versions, or where a revision is already mid-process.

    For anyone advising authors on authorship rights and responsibilities, the clearest guidance is to check funder licensing terms before first posting, since fixing a mismatched licence later depends on the preprint still being within the two-year eligibility window. Related open-research terminology, including licensing and reuse definitions, is tracked in the CASRAI open-research dictionary.

    Expect other preprint servers to face similar pressure as CC BY mandates spread across research funders. The direction of travel — author-initiated, platform-mediated licence correction rather than manuscript resubmission — is a practical template other repositories are likely to adopt as funder compliance checks tighten.

  • BioRxiv PubMed Indexing: How the NIH Pilot Works

    BioRxiv PubMed indexing is not automatic. Preprints reach PubMed through a single federal mechanism — the NIH Preprint Pilot, run by the U.S. National Library of Medicine (NLM) — which pulls in preprints that acknowledge direct NIH funding or carry an NIH-affiliated author, provided they were posted from 1 January 2023 onward under the pilot’s current phase.

    The NIH Preprint Pilot is an NLM programme that makes NIH-funded preprints from eligible servers — bioRxiv, medRxiv, arXiv, and Research Square — discoverable through PubMed Central (PMC) and PubMed ahead of formal peer review, with a corresponding citation added on a weekly cycle.

    What is the NIH Preprint Pilot?

    The NIH Preprint Pilot began in June 2020 as a narrow, COVID-19-only initiative. NLM made more than 3,300 preprints reporting NIH-supported SARS-CoV-2 research discoverable in PMC and PubMed between June 2020 and June 2022, testing whether preprint records could accelerate discovery during a public-health emergency.

    Phase 2 launched on 30 January 2023 and dropped the COVID-only restriction. It now covers any preprint that acknowledges direct NIH support and/or lists an NIH-affiliated author, posted to an eligible server on or after 1 January 2023. Eligible preprints are added to PMC on a weekly basis and receive a corresponding PubMed citation automatically — authors do not submit anything separately.

    How a preprint moves from bioRxiv to PubMed

    The pipeline is largely invisible to authors and runs on a fixed weekly cadence. NLM does not wait for a submission; it identifies eligible content and pulls it in automatically, then layers PubMed on top of the PMC record.

    • Identification: NLM text-mines new bioRxiv and medRxiv postings for NIH-support acknowledgements and cross-checks the NIH Office of Portfolio Analysis tool for NIH-affiliated authors.
    • PMC ingestion: Citation and abstract metadata are pulled from the preprint server’s machine-readable feed to build an “article header” record, and a PMCID is assigned immediately to enable rapid discovery.
    • PubMed record creation: Once the PMC record exists, NLM generates the corresponding PubMed citation the same week, tagged with publication type “Preprint.”
    • Full-text conversion: Preprints posted under a Creative Commons licence enter a separate workflow to produce archival full-text XML, a process NLM says takes a few days and enables full-text search within PMC.

    Every record carries a prominent yellow information panel confirming the work has not been peer-reviewed, and NLM runs weekly checks — against the bioRxiv API, the Crossref API, and the Europe PMC API — to link a preprint to its eventual journal version, updating the PubMed status to “Updated” once that link is confirmed.

    Which preprint servers qualify

    Only four servers currently feed the pilot. NLM evaluates candidate servers against a published checklist — clear non-peer-review labelling, transparent versioning, open licensing information, machine-readable metadata, and a public archiving policy — modelled on NIH’s 2017 interim-research-products guidance (NOT-OD-17-050) and COPE’s preprint discussion document.

    Server Subject scope Operator DOI registration
    bioRxiv Life sciences openRxiv (independent nonprofit, formerly a Cold Spring Harbor Laboratory service) Crossref
    medRxiv Health and clinical sciences openRxiv, with Yale University and BMJ as founding partners Crossref
    arXiv Physics, mathematics, computer science, quantitative biology Cornell University Crossref
    Research Square Multidisciplinary Research Square Company Crossref

    bioRxiv and medRxiv are the two servers most relevant to biomedical research administrators, since both fall under openRxiv, the independent nonprofit that took over operation of both platforms from Cold Spring Harbor Laboratory. openRxiv’s separation from a single host institution was framed explicitly around long-term sustainability for the two servers NIH now indexes directly — a governance detail that matters for anyone assessing the pilot’s durability, since NLM’s own eligibility criteria require a “publicly stated archiving strategy to ensure long-term access.”

    What this means for discoverability, DOIs, and citation

    PubMed indexing changes where a preprint can be found, not whether it can be cited. Every bioRxiv preprint already receives a DOI registered through Crossref at posting, which is what makes it part of the citable scientific record regardless of NIH eligibility.

    According to bioRxiv’s own FAQ, preprints are indexed by “Google, all other search engines, Google Scholar, Crossref, Semantic Scholar, Europe PubMed Central, and Preprint Citation Index (connected to the Web of Science)” independent of the NIH pilot — PubMed indexing is an additional, funder-gated channel layered on top of that baseline discoverability.

    One clarification worth making explicitly: bioRxiv and medRxiv do not carry a Scimago Journal Rank or an impact factor. Both metrics are journal-level indicators computed from peer-reviewed citation data; a preprint server is a distribution platform, not a journal, so no SJR score exists for bioRxiv as a whole, and any figure circulating under “bioRxiv impact factor” searches is not an NLM, Crossref, or Scimago-sourced metric.

    Indexing also does not substitute for compliance. NLM is explicit that even when a preprint sits in PMC under the pilot, the NIH Public Access Policy still requires the peer-reviewed, accepted author manuscript to be separately deposited via NIHMS, with its own PMCID reported as proof of compliance.

    Answer-first questions about bioRxiv and PubMed

    Does bioRxiv show up in PubMed?

    Yes, but only conditionally. A bioRxiv preprint appears in PubMed only if it acknowledges direct NIH funding or lists an NIH-affiliated author and was posted under Phase 2 of the NIH Preprint Pilot (from 1 January 2023). Non-NIH preprints stay discoverable via Google Scholar, Crossref, and Europe PMC instead.

    What is a preprint in PubMed?

    In PubMed, a preprint is a record carrying the publication type “Preprint,” which separates it from peer-reviewed literature in search filters. It displays a yellow information panel stating the work has not undergone peer review, and PubMed links it automatically to the journal version once one is published.

    Does bioRxiv count as published?

    No. bioRxiv distributes complete but unpublished manuscripts, so posting there is not equivalent to journal publication. A preprint carries a DOI and is part of the citable record, but it lacks the peer-review certification that ICMJE and COPE norms attach to a published article.

    Is it okay to cite bioRxiv?

    Yes. bioRxiv preprints receive a DOI through Crossref, making them formally citable, and are indexed by Google Scholar, Crossref, Semantic Scholar, and Europe PMC. Authors citing them should flag that the underlying findings have not yet completed peer review.

    Why other funders are watching the pilot

    NIH’s approach is unusual because it is infrastructural rather than a mandate: it does not require authors to preprint, it simply makes eligible preprints easier to find once posted. That distinction is why other funders are studying it rather than replicating it wholesale.

    cOAlition S, the funder coalition behind Plan S, already treats preprints as an acceptable route to satisfying immediate open-access requirements, but no cOAlition S member currently operates an equivalent centralised indexing pipeline into a national biomedical database. UKRI’s open access policy similarly recognises preprints as compliant interim outputs without building comparable PMC-style ingestion.

    For research administrators, the practical takeaway is that discoverability infrastructure and funder mandates remain two separate policy levers. NIH has built the first at meaningful scale; whether other national funders follow with their own PMC-equivalent indexing pipeline — rather than policy language alone — is the open question institutions tracking preprint compliance should watch through 2026 and beyond.

  • UKRI Funding Cuts: ODA Impact for Grant Holders

    UKRI’s Official Development Assistance (ODA) budget has been cut twice since 2021 — first when the UK reduced ODA from 0.7% to 0.5% of gross national income, then again from February 2025, when government set a further target of 0.3% of GNI by 2027. For grant holders, these UKRI funding cuts mean tighter ODA-eligible budgets, reprofiled projects and sharper competition for what remains of the international development research portfolio.

    Official Development Assistance is UK government aid spending that, under the OECD Development Assistance Committee framework the UK follows, must primarily promote the economic development and welfare of countries on the OECD DAC list of ODA recipients. UKRI channels a portion of its research and innovation budget through ODA-eligible programmes such as the Global Challenges Research Fund (GCRF) and the International Science Partnerships Fund (ISPF), and this ODA-specific funding line has been reduced independently of — and more severely than — UKRI’s core domestic science budget.

    What Are the UKRI ODA Funding Cuts?

    The UKRI ODA funding cuts refer to two distinct reductions to the ring-fenced aid spending that UKRI channels into development-focused research. The first came in February 2021, when the UK government cut total ODA spending from 0.7% to 0.5% of GNI, citing the fiscal impact of the Covid-19 pandemic. The second came in February 2025, when the government set a further reduction path to 0.3% of GNI by 2027, this time to fund an increase in defence spending.

    These cuts sit apart from the wider 2026 restructuring of UKRI’s £8 billion domestic research and innovation budget — the grant pauses, “always open” transitions and STFC savings programme already covered in general UKRI funding-cuts reporting. ODA is a distinct, ring-fenced funding line governed by international aid rules, not by UKRI’s strategic prioritisation exercise, so it is cut and managed on a different timetable by a different part of government — the Treasury and the Foreign, Commonwealth & Development Office (FCDO) — than UKRI’s domestic “buckets.”

    How Did the 2021 ODA Reduction Affect Grant Holders?

    The 2021 cut hit UKRI’s ODA portfolio disproportionately hard. According to written evidence submitted to the House of Commons Science, Innovation and Technology Committee, UKRI received £422 million in ODA funding in 2020-21, a year-on-year reduction of roughly 70% — considerably steeper than the roughly 50% cut (from £1.4 billion to £706 million) applied to the then Department for Business, Energy and Industrial Strategy’s ODA budget over the same period.

    UKRI responded by offering grant holders on GCRF and Newton Fund awards a limited set of options:

    • Termination of the award where the project could not be meaningfully continued at a reduced scale.
    • Reprofiling, adjusting project scope, partners or timelines to fit a smaller budget.
    • A case for exceptional funding, submitted where neither termination nor reprofiling was workable.

    UKRI also lifted virement restrictions across GCRF and Newton Fund grants, giving award holders flexibility to move funds between budget lines to protect priority activities. Grant holder surveys at the time found that most UK-based and in-country respondents expected the cuts to harm the communities the research was meant to benefit, and sector bodies including the Development Studies Association called for international partners to receive the same contractual protection as UK-based researchers.

    What Do the 2025–2027 ODA Reductions Mean for UKRI Grants?

    The 2025 announcement is a second, separate contraction. The government’s decision to reduce total UK ODA spending to 0.3% of GNI by 2027 was made to help fund a rise in defence spending, and it lands on top of a portfolio that had never fully recovered from the 2021 cut. UKRI published its own ODA review on 14 May 2025, examining options to mitigate the impact of the reduced allocation on its remaining GCRF and Newton Fund commitments.

    Reporting on the FCDO’s published ODA allocations for 2026/27 to 2028/29 indicates that bilateral aid — the category most research partnerships fall under — is being cut more heavily than multilateral aid, with Africa and the Middle East and North Africa region facing the largest reductions, and humanitarian aid, global health and education programmes among the hardest-hit sectors. Because UKRI’s development-research spending sits within the wider bilateral ODA envelope, these sector and regional priorities directly shape which UKRI-funded partnerships are likely to be protected or reduced.

    Milestone UK ODA target Effect on UKRI’s ODA-eligible research
    Pre-2021 baseline 0.7% of GNI GCRF and Newton Fund at planned commitment levels
    February 2021 0.5% of GNI UKRI’s ODA budget fell roughly 70% year-on-year; termination, reprofiling and exceptional-funding options offered
    14 May 2025 0.5% of GNI (interim) UKRI ODA review published to plan for further reductions
    By 2027 0.3% of GNI (target) Bilateral ODA facing the steepest cuts; sharper competition for renewal and new ODA-eligible awards

    What Should Current and Prospective ODA Grant Holders Do Now?

    Grant holders with live or planned UKRI ODA awards should treat the funding line as structurally constrained through at least 2027, not as a temporary pause. Practical steps include:

    1. Confirm commitment status directly with your UKRI council rather than assuming an existing award is automatically protected — legal commitments for prior awards are generally honoured, but renewal and extension decisions are being made against a shrinking envelope.
    2. Build reprofiling flexibility into proposals, mirroring the virement and scope-adjustment approach UKRI used during the 2021 cuts, so a project can absorb a mid-award budget reduction without collapsing.
    3. Prioritise equitable partner treatment, applying a “do no harm” approach that protects in-country and Global South partners’ contracts and payments on the same terms as UK-based staff.
    4. Diversify beyond a single ODA-eligible instrument, exploring co-funding through Horizon Europe association, the International Science Partnerships Fund, or bilateral country-level schemes where GCRF or Newton Fund successors have narrowed.
    5. Track FCDO sector and regional priorities, since bilateral ODA cuts are not applied evenly — health, education and humanitarian-linked research in Africa and the MENA region face the steepest reductions.

    Answer-First Q&A: UKRI ODA Funding Cuts

    Is UKRI ODA funding being cut again?

    Yes. UKRI’s Official Development Assistance funding remains under sustained pressure following the UK government’s February 2025 decision to reduce total ODA spending to 0.3% of gross national income by 2027. UKRI published an ODA review in May 2025 to plan how this affects its Global Challenges Research Fund and Newton Fund portfolios.

    What is ODA funding at UKRI?

    Official Development Assistance is UK government aid spending that must promote the economic development and welfare of low- and middle-income countries. UKRI channels ODA through programmes including the Global Challenges Research Fund and the International Science Partnerships Fund, subject to strict eligibility rules and Researchfish outcome reporting.

    Will UKRI honour existing ODA grant commitments?

    UKRI has generally protected legal commitments for already-awarded projects, but new and renewal ODA awards face tighter competition as the overall envelope shrinks. Grant holders should confirm commitment status directly with their UKRI council rather than assume automatic continuity through to project completion.

    How is ODA funding different from UKRI’s core science budget?

    ODA funding is ring-fenced aid spending tied to development outcomes in eligible countries, separate from UKRI’s domestic research and innovation budget, which is rising toward almost £10 billion annually by 2030. Cuts to the aid budget do not automatically reduce curiosity-driven domestic grants, but they do shrink international development-focused funding specifically.

    Implications and What Happens Next

    For institutions, the practical implication is that ODA-eligible research administration now needs its own risk register, separate from general UKRI grant-pipeline planning. Research offices should flag ODA-funded projects for early review, model reprofiling scenarios before a reduction is announced rather than after, and ensure partnership agreements with Global South institutions include contingency clauses for budget reductions.

    Sector bodies continue to advocate for a return to the 0.7% ODA commitment, but the government’s 2025 decision was framed as a multi-year fiscal choice tied to defence spending, not a temporary measure. Grant holders and research administrators managing UKRI ODA awards should plan on a constrained funding line running through the current spending review period to 2027, with FCDO bilateral sector and regional priorities — not UKRI’s domestic strategic “buckets” — determining which development-research partnerships come under the most pressure next.

  • MRC Funding Update UKRI: Grants Reopened

    MRC funding update, in brief: the Medical Research Council paused several applicant-led grant schemes from February 2026 while UK Research and Innovation restructured to an “always open” application model; research grants, new investigator grants and partnership grants reopened on 7 April 2026, experimental medicine opportunities reopened on 30 April 2026, and MRC Proof of Concept and Impact Acceleration Awards are scheduled to reopen in July 2026. For grant holders, the practical implications are a new rolling submission window, a consolidated review structure, and firm caps on how many applications one person can lead at once.

    The MRC funding update ukri has published since February 2026 marks the most significant procedural change to Medical Research Council grant administration in over a decade. Medical Research Council (MRC) is one of UK Research and Innovation’s seven disciplinary councils, responsible for funding biomedical and health research across the UK’s higher education and institute sector. This briefing sets out exactly what changed, what remains open, and what research administrators and principal investigators need to do differently when planning 2026 applications.

    What actually changed in the MRC funding update

    On 1 February 2026, UKRI Chief Executive Ian Chapman issued an open letter to the research and innovation community announcing a new investment approach for the 2026–2030 spending review period. The letter described a shift toward a “more strategic, UKRI-wide model” for funding decisions, and confirmed that MRC would use the transition to refresh its approach to applicant-led funding.

    The headline structural change is a move to an “always open” responsive-mode system, replacing fixed application deadlines. UKRI states that published deadlines “cause significant variation in the volume of applications we receive and in reviewer availability,” and that removing them smooths these peaks and troughs. This is an operating-model change, not a funding cut — MRC says curiosity-driven research remains a committed priority, underpinned by a UKRI-wide 50% budget commitment to that category of research.

    To implement the “always open” system, MRC had to pause several applicant-led funding opportunities from February 2026 while the assessment infrastructure was rebuilt. Awards that had already been offered, accepted or started were explicitly unaffected throughout the transition.

    Reopening timeline: what’s open now and what’s still paused

    As of the most recent UKRI update (15 June 2026), most MRC applicant-led schemes have reopened. Grant holders should treat the table below as the operative reference, not the earlier February pause notice, which is now superseded.

    MRC funding opportunity Status Reopening date
    Applicant-led research grants Open 7 April 2026
    New investigator research grants Open 7 April 2026
    Partnership grants (applicant-led) Open 7 April 2026
    Experimental medicine opportunities Open 30 April 2026
    MRC Proof of Concept (formerly Developmental Pathway Funding Scheme, stage one) Reopening July 2026
    MRC Impact Acceleration Awards (formerly MRC Gap Fund) Reopening July 2026
    Fellowships, studentships, Centres of Research Excellence Never paused Continuously open

    UKRI expects the wider transition to be complete by the start of the 2027–2028 financial year (6 April 2027 to 5 April 2028). Institutions running internal peer-review or costing pipelines timed to the old deadline calendar should recalibrate now: under “always open” mode, there is no annual cycle to plan around.

    The new College of Experts review structure

    MRC’s four disciplinary research boards — covering infections and immunity, molecular and cellular medicine, neurosciences and mental health, and population and systems medicine — have been consolidated into a single College of Experts. Funding panels are now drawn flexibly from this combined pool rather than fixed to a single board.

    This restructuring supports cross-disciplinary applications that previously sat awkwardly between boards, and enables faster decisions by decoupling panel composition from a rigid quarterly schedule. Applications closed before the transition — including the legacy boards’ November 2025 round and Developmental Pathway Funding Scheme stage two — are still assessed under the old structure, with decisions expected in April 2026; MRC has confirmed a reduced number of awards from that backlog, reflecting the changeover rather than any change in typical grant size going forward.

    Application caps and resubmission rules grant holders must know

    Two new eligibility mechanics apply under the “always open” model and directly affect how principal investigators should sequence their applications:

    • Application cap: a maximum of two applications as project lead may be submitted across applicant-led responsive-mode funding calls within any rolling 12-month period.
    • Resubmission bar: an application previously unsuccessful with MRC — or with any other funder — will not be considered again for 12 months, unless MRC has explicitly invited a resubmission.
    • Cost basis unchanged: MRC continues to fund 80% of the full economic cost (FEC), with grant durations ranging from 18 months to five years and no fixed cap on requested amount, provided the sum is proportionate to project scope.

    For research offices, “always open” does not mean unlimited throughput per investigator — it shifts the constraint from a calendar deadline to a rolling personal quota. Grant-writing capacity planning built around a fixed autumn or spring deadline now needs continuous tracking of each investigator’s rolling 12-month application count.

    Budget signals: what the wider UKRI settlement means

    UKRI has stated that its overall research and innovation budget is rising across the 2026–2030 spending review period, and that the budget for biomedical and health research specifically is “in an excellent position.” Independent analysis from the Campaign for Science and Engineering notes that the overall UKRI budget is set to rise toward £10 billion a year by 2030, though how that funding is distributed across councils and themes is shifting considerably as part of the same restructuring.

    A parallel strand channels additional funding through the UKRI Life Sciences Priority Programme — a cross-council theme through which MRC accesses coordinated funding beyond its standalone curiosity-driven allocation. UKRI frames this as additive: fellowships, studentships and Centres of Research Excellence funding was unaffected throughout, and the pause applied only to specific applicant-led schemes during infrastructure changes.

    BBSRC underwent the same “always open” transition in parallel; its new investigator award and standard research grant have also reopened. Grant holders working across MRC and BBSRC funding lines should expect the same rolling-quota and resubmission mechanics on both councils, since the operating model is shared across the UKRI-wide transition rather than council-specific.

    Frequently asked questions

    Why did MRC pause its funding opportunities in 2026?

    MRC paused several applicant-led schemes from February 2026 to implement UKRI’s move to an “always open” application system. UKRI stated that fixed deadlines caused uneven application volumes and reviewer availability problems, and that removing them required behind-the-scenes changes to assessment infrastructure before reopening.

    Which MRC grants have reopened?

    Applicant-led research grants, new investigator research grants and partnership grants reopened on 7 April 2026, and experimental medicine opportunities reopened on 30 April 2026. MRC Proof of Concept and MRC Impact Acceleration Awards are scheduled to reopen in July 2026.

    Were MRC fellowships and studentships affected?

    No. MRC has confirmed that funding for fellowships, studentships, and MRC Centres of Research Excellence was not affected by the 2026 pause and remained continuously open for applications throughout the transition period.

    How many MRC applications can one person lead at once?

    Under the new rules, a principal investigator may lead a maximum of two applications across applicant-led responsive-mode funding calls within any rolling 12-month period, and unsuccessful applications face a 12-month resubmission bar unless MRC has invited one.

    Implications for institutions and grant holders

    Research offices should update three things now. First, replace deadline-driven internal sign-off calendars with continuous submission tracking, since “always open” removes the predictable peaks institutions have historically planned costing and QA cycles around. Second, build a per-investigator rolling application count into grants-management systems to enforce the two-applications-per-12-months cap before a proposal reaches MRC and is rejected on eligibility grounds. Third, brief investigators explicitly on the 12-month resubmission bar — a previously unsuccessful proposal, even a strong one, is not eligible for quick resubmission without an explicit MRC invitation, which changes revision strategy considerably.

    MRC has said it will continue to share updates and reopen remaining funding opportunities as they become ready, with full transition to the new model expected by April 2027. Institutions with active or upcoming submissions should monitor the MRC application timeline directly rather than relying on the February 2026 pause notice, which the June 2026 update has substantially superseded.

  • Research Misconduct Policy: Lehigh’s 2026 Update Sets the Pace for Peer Institutions

    A research misconduct policy is the institutional and federal framework that defines fabrication, falsification, and plagiarism (FFP) in research, and sets out how allegations are reported, investigated, and adjudicated. Lehigh University’s new research misconduct policy took effect on 1 January 2026, and it is not an isolated update: dozens of US research universities revised their policies for the same date, all responding to the first substantial rewrite of the federal research misconduct rule in two decades.

    Research misconduct policy in the United States is not institutional discretion. Any university receiving Public Health Service (PHS) funding — the great majority of federally funded research institutions — must maintain a policy satisfying 42 CFR Part 93, administered by the HHS Office of Research Integrity (ORI). When that regulation changes, every covered institution’s policy changes with it. That happened in 2024, and 1 January 2026 is when the consequences arrived on campus.

    What does Lehigh’s new research misconduct policy change?

    Lehigh’s Office of Research Integrity finalised its revised policy on 19 December 2025, in a notice from Naomi E. Coll, Director of Research Policy and Compliance, confirming that the policy would take effect on 1 January 2026. The document sets out how the university assesses, inquires into, and investigates allegations of research misconduct against academic, scientific, and professional staff, students, and contractors, regardless of funding source.

    What distinguishes Lehigh’s process is its provenance rather than its content: the policy was drafted by a faculty working group with one representative from each college, presented to the Faculty Senate in November 2025, and opened to a university-wide online public comment period before finalisation. That sequence — draft, Senate review, public comment, publication — is now a template other institutions are following as they race the same federal deadline.

    What is the federal research misconduct policy institutions must follow?

    The federal research misconduct policy originates with the Office of Science and Technology Policy’s 2000 Federal Policy on Research Misconduct and is codified for PHS-funded research at 42 CFR Part 93. It defines research misconduct as fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results — honest error and honest differences of opinion are explicitly excluded.

    A formal finding requires three elements together: a significant departure from accepted practices of the relevant research community, conduct that was intentional, knowing, or reckless, and proof by a preponderance of the evidence. Non-PHS agencies apply parallel rules — the National Science Foundation administers its own version at 45 CFR Part 689 — so multi-funder institutions must reconcile more than one regulatory text in a single policy.

    What changed in the 2024 ORI Final Rule (42 CFR Part 93)?

    ORI published its final rule in the Federal Register on 17 September 2024 — the first substantive revision to the 2005 regulation. The rule became legally effective on 1 January 2025, but institutions only had to apply it to allegations received on or after 1 January 2026, which is why the current wave of policy rewrites is landing now rather than a year ago.

    • Defined intent standard: “Intentionally,” “knowingly,” and “recklessly” — previously undefined terms central to any finding — now carry fixed regulatory definitions.
    • New “institutional record” definition: the rule specifies, for the first time, exactly which documents (inquiry and investigation reports, sequestered evidence indexes, appeal records) must be compiled and preserved.
    • Seven-year retention rule: institutions must secure the institutional record and all sequestered evidence for seven years after the proceeding — or any related HHS proceeding — concludes, whichever is later.
    • Longer inquiry window: the initial inquiry phase is extended from 60 to 90 days, giving Research Integrity Officers more time before deciding whether a full investigation is warranted.

    ORI has confirmed it will not require institutions to submit revised policies until the annual report covering 2025 is filed, due by 30 April 2026 — but the policies themselves, and the procedures behind them, must already be in force for any allegation received from 1 January 2026 onward.

    Which other universities are revising policies for 2026?

    Lehigh is one of a cohort. Institutional policy pages and faculty handbooks show a cluster of revisions timed to the same compliance date, though each institution has followed its own governance path to get there.

    Institution Policy action Date Notes
    Lehigh University New research misconduct policy finalised Effective 1 Jan 2026 Faculty working group, Senate review, public comment
    Columbia University Revised policy approved by University Senate Approved 12 Dec 2025; effective 1 Jan 2026 Realigns definitions with the PHS Final Rule
    Auburn University Policy on Research Misconduct republished 21 Nov 2025 Updated allegation-resolution framework
    Ohio State University Research Misconduct policy revised March 2025 Among the earliest movers ahead of the deadline
    Wayne State University Interim Research Misconduct Policy (26-01) Interim, 2025–26 cycle Document numbering signals the 2026 compliance cycle
    Duke University Misconduct in Research policy (established) Bound to 42 CFR Part 93 / 45 CFR Part 689 Long-cited model policy; must reflect the revised rule by the deadline

    The pattern is consistent: governance bodies — Faculty Senate, University Senate, or equivalent — are being used explicitly to legitimise the revision, not just administrative sign-off. That marks a shift from prior cycles, where research misconduct policy updates were typically handled as compliance housekeeping rather than faculty-governance business.

    Common questions on research misconduct policy

    What are the three forms of research misconduct under US federal policy?

    US federal policy recognises three forms of research misconduct: fabrication (making up data or results), falsification (manipulating materials, equipment, or data so the research record is inaccurate), and plagiarism (appropriating another person’s ideas, processes, results, or words without credit). This “FFP” definition has applied since the 2000 OSTP policy and is unchanged by the 2026 rule.

    What is considered research misconduct in the US?

    Conduct only counts as research misconduct if it is a significant departure from accepted practices of the relevant research community, committed intentionally, knowingly, or recklessly, and proven by a preponderance of the evidence. Honest error, differences of scientific opinion, and most authorship disputes are explicitly excluded from the definition.

    What is 42 CFR Part 93?

    42 CFR Part 93 is the codified federal regulation — the Public Health Service Policies on Research Misconduct — that sets binding rules for how PHS-funded institutions must respond to allegations. Its 2024 revision, effective for allegations from 1 January 2026, is the first major overhaul since 2005 and drives the current round of university policy rewrites.

    What does the federal research misconduct policy cover?

    The federal research misconduct policy covers federally funded research and funding proposals across all fields — including medicine, social sciences, engineering, and statistics — conducted at universities, federal laboratories, and other funded research institutions. It governs the inquiry, investigation, and adjudication phases but does not supersede separate rules on human-subjects protection, animal welfare, or financial conflicts of interest.

    What this means for research administrators

    The consequences of research misconduct under the federal framework range from correction of the research record and letters of reprimand to suspension of funding, debarment from future federal awards, and referral for civil or criminal fraud proceedings. For research administrators, the sharper immediate risk is procedural: if an institution’s inquiry, evidentiary, or retention procedures do not match the 2026 rule, findings made under a defective process are vulnerable on appeal.

    Three takeaways follow. First, Research Integrity Officers should confirm their policy explicitly defines “intentionally,” “knowingly,” and “recklessly” in the regulatory terms now fixed by the final rule, rather than leaving them undefined as most pre-2026 policies did. Second, retention schedules need extending to the mandated seven years and mapped against the newly defined “institutional record.” Third, institutions routing revisions through faculty governance — as Lehigh, Columbia, and Auburn have done — are building a stronger due-process record that may matter if a finding is later contested.

    The compliance clock does not stop at 1 January. ORI’s own review of revised policies, via the annual report covering 2025, falls due on 30 April 2026 — a second checkpoint on which institutions treated this as a genuine revision to research administration practice, and which merely updated a document to survive it.

  • Wellcome Trust Open Access Policy Explained: Requirements, APC Funding and the Plan S Link

    Wellcome Trust’s open access policy requires that original research articles arising, in whole or in part, from Wellcome funding be made freely available in Europe PMC immediately on publication, licensed CC BY, with no embargo permitted — a rule that has applied to all qualifying submissions since 1 January 2021 and is backed by direct grant-linked funding for publication costs.

    The Wellcome Trust open access policy is the funder’s mandatory requirement that peer-reviewed research articles, monographs and book chapters supported by its grants be deposited in Europe PMC (and NCBI Bookshelf, for long-form works) and openly licensed at the point of publication, with compliance routes and funding administered directly by Wellcome rather than through a separate national mandate.

    What Wellcome’s open access policy requires

    Wellcome’s current policy took effect for articles submitted from 1 January 2021. It applies to all original peer-reviewed research articles supported wholly or partly by Wellcome funding, and requires immediate, unembargoed open access with a CC BY licence (CC BY-ND only by case-by-case exception, requested before submission).

    Three routes satisfy the policy:

    • Gold route — publish in a fully open access journal or platform; the publisher deposits the version of record in PubMed Central and Europe PMC.
    • Green route — publish in a subscription journal and make the author accepted manuscript (AAM) open in Europe PMC immediately, under CC BY, via Wellcome’s rights retention clause.
    • Transformative-agreement route — publish through an institutional transformative agreement that makes the version of record open access.

    Scholarly monographs and book chapters carry a separate provision: they may be deposited up to six months after publication and are not required to use CC BY if another Creative Commons licence is more appropriate.

    How Wellcome funds APC and open access costs

    Wellcome funds open access directly rather than relying solely on national block-grant schemes. In the 2021/22 grant year, Wellcome awarded £7.2 million in open access block grants to 38 institutions, and 92% of individually reported articles complied with the policy, according to Wellcome’s own published guidance.

    The funding rules tightened materially from 1 January 2025: Wellcome now funds article processing charges only for research articles published in fully open access journals or platforms. Hybrid (“paywall-plus-OA-option”) subscription journals are no longer eligible for Wellcome APC funding, even where an institution holds a transformative agreement with the publisher — a stricter position than Wellcome held during 2021–2024, when transformative-agreement “publish” fees were still fundable.

    This mirrors the broader Plan S trajectory away from hybrid subsidy, but Wellcome reached the fully-OA-only funding line on its own timetable, separate from any coalition-wide deadline.

    Rights retention: the automatic CC BY clause

    Wellcome’s rights retention mechanism is built into its grant conditions rather than delivered through a separate author addendum. Since 1 January 2021, Wellcome grant terms automatically apply a CC BY licence to the author accepted manuscript of any original research article the grant supports, in whole or in part — meaning the author retains sufficient rights to make the AAM open immediately, regardless of a publisher’s default embargo.

    This differs in mechanism, though not in intent, from UKRI’s parallel approach. UKRI’s own open access policy, which took effect for journal articles submitted from 1 April 2022, established a comparable zero-embargo, CC BY-on-AAM route roughly fifteen months after Wellcome’s clause had already been in force.

    Funder Zero-embargo Green route Automatic CC BY on AAM effective from cOAlition S / Plan S status
    Wellcome Trust Yes 1 January 2021 Founding member
    UKRI Yes 1 April 2022 Founding member
    cOAlition S baseline (Plan S) Required for Green route Recommended coalition-wide via the Rights Retention Strategy Coalition framework itself

    Where an author cannot secure rights retention on the AAM and the article is not published Gold, several UK institutions — including UCL — accept a CC BY-licensed preprint deposited in a Europe PMC-indexed preprint server before publication as an alternative compliance path.

    Wellcome is a founding member of cOAlition S, the funder consortium that launched Plan S in September 2018 alongside Science Europe and the European Commission, and that now counts roughly two dozen funder organisations among its members. Wellcome’s current policy is deliberately aligned with Plan S’s core principles: immediate access, open licensing, and author choice of venue.

    But the relationship runs the other way chronologically. Wellcome introduced one of the world’s first funder open access mandates in 2006 — over a decade before Plan S existed — requiring deposit of funded research in PubMed Central. The 2021 policy update did not create Wellcome’s open access commitment; it tightened an existing mandate (removing the embargo Wellcome had previously permitted) specifically to bring it into line with Plan S’s stricter, zero-embargo standard. Wellcome-funded researchers can check which compliance route a given journal supports using the Journal Checker Tool, built jointly by cOAlition S and its funder members.

    Common questions about Wellcome’s policy

    Does Wellcome Trust pay for open access publishing?

    Yes. Wellcome funds APCs directly for articles in fully open access journals and platforms, distributes annual open access block grants to grant-holding institutions, and separately funds compliant open access publication of monographs and book chapters on request.

    What licence does Wellcome require for open access articles?

    CC BY is the default requirement for all research articles made open under the policy. A more restrictive CC BY-ND licence is permitted only by prior, case-by-case exception, requested through Wellcome’s dedicated request form before submission.

    Is Wellcome Trust part of Plan S?

    Yes. Wellcome is a founding member of cOAlition S, the funder coalition behind Plan S. Its own open access policy is fully aligned with Plan S principles, though Wellcome’s underlying open access mandate predates Plan S by more than ten years.

    What happens if a Wellcome-funded paper doesn’t comply?

    Non-compliant articles are not automatically penalised per se, but Wellcome tracks compliance rates at institutional level, ties future open access block grant funding to institutional performance, and expects grant-holding organisations to actively support researchers in meeting the policy’s routes.

    Implications for institutions and researchers

    For research administrators managing Wellcome grant portfolios, the practical shift is the narrowing of fundable routes since January 2025: budgeting for hybrid-journal APCs against Wellcome funds is no longer viable outside a qualifying transformative agreement’s non-hybrid terms, so pre-submission journal checking has become a compliance necessity rather than a courtesy.

    Researchers publishing in subscription journals should treat rights retention as the default fallback, since it requires no publisher fee and no embargo negotiation — the CC BY licence on the accepted manuscript is already secured by the grant terms before the paper is ever submitted.

    Outlook

    Wellcome’s trajectory — mandate first, coalition alignment second, funding restriction third — has become a template other funders are following at their own pace. As UKRI, NIHR and other cOAlition S members continue tightening hybrid-journal funding eligibility, institutions that already built Wellcome-compliant workflows around zero-embargo Green routes are better placed to absorb the next round of funder-specific restrictions.

  • Gates Foundation Open Access Policy: No More APCs for cOAlition S Funders

    The Gates Foundation open access policy was refreshed for 2025, taking effect on 1 January 2025: the foundation stopped paying article processing charges (APCs) for individual manuscripts, added a mandatory preprint-deposit requirement, and expanded the policy’s scope to cover every funded manuscript and its underlying data.

    The Gates Foundation open access policy is the Bill & Melinda Gates Foundation’s mandatory framework requiring that all peer-reviewed research and data arising from its funding be made freely available, openly licensed, and reusable without embargo. As a founding member of cOAlition S, the funder that co-created and popularised Plan S in 2018, the foundation’s 2025 refresh is being watched closely as a signal of where other research funders may be heading on APC costs.

    What Changed in the 2025 Policy Refresh?

    The 2025 policy is framed as a “refresh” of the 2021 policy, not a wholesale replacement — the core repository mandate survives intact. What changes is scope and support. It now applies to “all published research funded, in whole or in part, by the foundation” — termed Funded Manuscripts — and to “any data underlying the Funded Manuscripts,” a broader remit than the 2021 text.

    Three elements define the refresh:

    • Mandatory preprint deposit for every funded manuscript, in addition to the existing accepted-manuscript deposit requirement.
    • An earlier trigger for open data: data must now be accessible as soon as the preprint is available, not only once the accepted article is published.
    • Withdrawal of financial support for individual APC payments, shifting that cost onto grantees and co-authors.

    Why Did the Foundation Stop Paying APCs?

    The 2025 policy is unambiguous on this point: “The Foundation Will Not Pay Article Processing Charges (APC). Any publication fees are the responsibility of the grantees and their co-authors.” This is a change to support, not to what is mandated — grantees were never obliged to use the foundation’s APC funds under the 2021 policy, but many did, particularly for publishing in fully open-access journals listed in the Directory of Open Access Journals (DOAJ).

    The foundation frames this as part of a wider push against the gold-OA/APC model, which critics argue rewards well-resourced authors while pricing out others. Rather than underwriting per-article fees, it says it will back non-APC routes to open publishing, including its own Gates Open Research platform and select publisher partnerships covering Gates-funded authors outside the standard APC mechanism.

    The practical effect falls unevenly. Per an analysis by Lisa Janicke Hinchliffe in The Scholarly Kitchen (15 April 2024), Gates grantees publish roughly 4,000 papers a year — about 0.07% of articles published globally — so the aggregate revenue impact on any single open-access publisher is likely modest, even though the effect on individual grantees lacking alternative funding can be significant.

    What Does the New Preprint-First Mandate Require?

    Every Funded Manuscript must now be “published as a preprint in a preprint server recognized by the foundation” that applies sufficient scrutiny, carrying a CC BY 4.0 licence or equivalent. This sits alongside — not instead of — the existing requirement that the accepted manuscript be deposited “immediately upon publication in PubMed Central (PMC), or in another openly accessible repository, with proper metadata tagging identifying Gates funding.”

    Two details matter for compliance teams:

    • Grantees can self-exempt from the preprint requirement where they determine “a preprint is not appropriate due to ethical, safety or other legitimate concerns” — the foundation has not yet published criteria for what counts.
    • The foundation is not mandating a single preprint server. It has said it will point grantees to ASAPbio’s preprint server directory rather than maintain its own list, though it separately partnered with Taylor & Francis/F1000 to launch VeriXiv, a verified preprint platform grantees may optionally use.

    Copyright-retention language is essentially unchanged: grantees must retain enough copyright to deposit and licence the manuscript CC BY 4.0, and include a foundation-mandated acknowledgement and rights-retention statement.

    How Does Gates Compare With Other cOAlition S Funders?

    The Gates Foundation is not the only cOAlition S member re-examining its terms, but it is the first founding funder to formally withdraw central APC funding. Wellcome Trust and UK Research and Innovation (UKRI), the two other funders most closely associated with Plan S’s origins, still fund publication fees through institutional block-grant mechanisms rather than the pay-per-article support the Gates Foundation has now dropped.

    Funder Current policy effective date Pays APCs? Preprint requirement
    Gates Foundation 1 January 2025 (refresh) No — discontinued for individual manuscripts Mandatory, narrow ethical/safety exemption
    Wellcome Trust 1 January 2021 Yes — via institutional block grants Encouraged, not mandated
    UKRI 1 April 2022 (journal articles); 1 January 2024 (monographs) Yes — via block grant to research organisations Not mandated
    cOAlition S / Plan S baseline Founding principles from 2018 Funder-dependent; no central APC cap post-2024 Not centrally mandated; Rights Retention Strategy supported

    cOAlition S has been diplomatic about the change. Responding to the refresh, Executive Director Johan Rooryck said: “Five years on since Plan S was first published, it is entirely appropriate that funders are reviewing their OA policies to ensure they are effectively meeting their goals… Our collective dedication to making full and immediate OA a reality remains the driving force behind our collaboration.” Separately, Rooryck told Nature the refreshed policy is not “entirely in line” with cOAlition S guidance, while noting member funders retain “a lot of leeway” in how they implement shared principles. The Scholarly Kitchen nonetheless judges the policy relatively well aligned with Plan S’s repository route, since the accepted-manuscript deposit mandate — the mechanism that satisfies Plan S compliance — is retained, not replaced.

    Common Questions About the Gates Foundation Open Access Policy

    Does the Gates Foundation still pay APCs?

    No. Since the 2025 Open Access Policy took effect on 1 January 2025, the Bill & Melinda Gates Foundation no longer pays article processing charges for individual manuscripts. Publication fees are now the responsibility of grantees and co-authors, though the foundation continues to fund open infrastructure and select publisher arrangements.

    Does the Gates Foundation require preprints?

    Yes. Under the 2025 policy, every funded manuscript must be posted as a preprint on a foundation-recognised server carrying a CC BY 4.0 licence, in addition to the existing requirement to deposit the accepted manuscript in PubMed Central or another open repository. Grantees may seek an exemption only for documented ethical or safety concerns.

    Is the Gates Foundation still aligned with Plan S?

    Largely, yes. cOAlition S publicly welcomed the 2025 refresh, and analysts judge the policy broadly consistent with Plan S‘s repository route despite the APC-funding withdrawal. Executive Director Johan Rooryck said the update reflects funders “reviewing their OA policies,” while stopping short of declaring it fully in line with coalition guidance.

    What are the Gates Foundation’s data-sharing requirements?

    The 2025 policy requires underlying data to be openly accessible immediately once the funded manuscript becomes available — including at the preprint stage. This is earlier than the 2021 policy, which triggered the data-sharing mandate only once the accepted article was formally published in a journal.

    Implications and What to Watch Next

    For grantees without alternative funding, the squeeze is real: authors who relied on Gates APC support for DOAJ-listed open-access journals must now find a fee waiver, an institutional agreement, or a non-APC venue, while still meeting deposit and rights-retention requirements some publishers only accommodate via paid gold-OA routes. Research administrators managing multi-funder compliance will need to track this alongside UKRI and Wellcome Trust obligations, since the three funders no longer follow a uniform APC-support model despite shared Plan S origins.

    For publishers, immediate revenue exposure looks limited given the modest volume of Gates-funded output, but the policy adds pressure toward non-APC business models — waiver programmes, “pure publish” institutional agreements, and preprint-native platforms — that cOAlition S’s own “Beyond Article-Based Charges” working group, established with Jisc and PLOS, is separately examining.

    The signal for other funders is the more consequential story. Gates is the first cOAlition S founding member to formally withdraw central APC funding while retaining a Plan S-compatible repository mandate. Whether Wellcome Trust, UKRI, or other coalition funders follow with funding recalibrations — rather than eligibility or embargo changes — is the development worth monitoring as institutions plan multi-year compliance budgets.

    Research administrators managing multi-funder open access compliance can find related standards context in CASRAI’s research administration resources.

  • AI Regulations Around the World: A 4-Jurisdiction Comparison for Research Consortia

    AI regulations around the world diverge sharply in 2026: the EU AI Act is a binding, risk-tiered statute with extraterritorial reach; US oversight is a fragmented state-law patchwork with no federal statute; China runs a registration-and-content-control regime; and the Council of Europe treaty sets shared principles without direct enforcement. No single checklist covers every consortium partner.

    AI regulation is the set of binding statutes, administrative measures and international instruments that govern how artificial intelligence systems may be developed, deployed and used within a given jurisdiction. For a multi-country research consortium, that matters practically: the same AI-assisted analysis tool or generative writing aid can be lawful for one partner and non-compliant for another, purely because of where each institution sits.

    This article maps the four regimes that most often collide in international data-sharing agreements — the EU AI Act, US state AI laws, China’s AI measures, and the Council of Europe’s AI treaty — and identifies where the friction actually occurs, not just what each law says in isolation.

    Contents

    What are the main AI regulatory regimes research consortia must track?

    Four regimes dominate cross-border research collaboration in 2026: the EU’s binding AI Act, a growing set of US state statutes, China’s registration-led administrative measures, and the Council of Europe’s rights-based treaty. Each uses a different legal mechanism, a different geographic trigger, and a different enforcement model, which is precisely why a consortium cannot rely on one partner’s compliance work to cover the group.

    Regime Legal status Geographic reach Compliance approach Relevance to research consortia
    EU AI Act (Regulation (EU) 2024/1689) Binding regulation, directly applicable in all EU member states Extraterritorial: covers providers and deployers outside the EU where an AI system’s output is used within the EU Risk-tiered (unacceptable / high / limited / minimal); high-risk obligations broadly apply from 2 August 2026 A non-EU partner can still be caught if any consortium member deploys the tool’s output in the EU; grant-review and admissions-style AI can fall into Annex III high-risk categories
    US state AI laws No binding federal AI statute; state statutes such as the Colorado AI Act and California’s AI Transparency Act Applies within the enacting state; a December 2025 executive order pushes federal preemption of “burdensome” state rules, but this is contested and unsettled as of mid-2026 Sector- and harm-specific (algorithmic discrimination, transparency, deepfakes) rather than one risk taxonomy A single US institution can trigger several inconsistent duties depending on which state its staff, servers or subcontractors sit in
    China’s AI measures Binding administrative measures enforced by the Cyberspace Administration of China (CAC), now folded into the amended Cybersecurity Law from January 2026 Applies to AI services offered within China; requires algorithm registration with the CAC before deployment Registration- and content-control-led: mandatory labelling of AI-generated content, security assessments, real-name verification Chinese partner institutions typically cannot lawfully run an unregistered foreign AI tool against shared data, creating a hard blocker for joint analysis pipelines
    Council of Europe AI treaty First legally binding international AI treaty (Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law) Open to Council of Europe members and non-member signatories, including the US, UK, Canada and Japan; needs ratification by five signatories, including three Council of Europe states, to enter into force Principles-based: human rights, democracy and rule-of-law safeguards for public- and private-sector AI Offers shared language consortia can cite in data-sharing agreements, but is not self-executing and needs domestic implementing law in each signatory state

    How does the EU AI Act apply to multi-country research?

    The EU AI Act entered into force on 1 August 2024, with obligations phased in over several years. Bans on unacceptable-risk systems and AI-literacy duties applied from 2 February 2025; general-purpose AI obligations followed on 2 August 2025; and high-risk system obligations broadly apply from 2 August 2026, with penalties reaching €35 million or 7% of global annual turnover.

    What consortium leads consistently underestimate is scope. Like the GDPR before it, the Act reaches beyond EU borders: it applies to providers and deployers established outside the Union whenever the output of their AI system is used within it. A US or Asia-based partner running an AI-assisted screening tool that feeds results into an EU-led work package can be pulled into EU obligations even without an EU office. Annex III’s high-risk categories — including systems used in education, employment and essential services — also reach some AI-assisted grant-review and research-integrity screening tools.

    What do US state AI laws mean for consortium partners?

    The United States has no comprehensive federal AI statute in 2026. Instead, regulation is set state by state: Colorado’s AI Act (SB 24-205), the first comprehensive US state AI law, requires reasonable care to prevent algorithmic discrimination in high-risk systems, with implementation delayed to 30 June 2026. California has separately enacted an AI Transparency Act and a frontier-model safety statute.

    A December 2025 executive order directed federal agencies to challenge state AI laws viewed as inconsistent with a lighter-touch national standard, but as of mid-2026 that preemption push is unsettled and existing state statutes remain in force. For a consortium, a single US institution’s obligations can shift depending on which state its staff, infrastructure or subcontractors sit in — and may change again if preemption litigation succeeds.

    How does China regulate AI differently?

    China’s approach is registration-led and content-focused rather than risk-tiered. The Cyberspace Administration of China requires algorithm registration and security assessment before many AI services can be deployed. The Measures for Labelling AI-Generated and Synthetic Content took effect in September 2025, three national standards on generative AI security took effect on 1 November 2025, and AI governance obligations were folded into the amended Cybersecurity Law from January 2026.

    For research consortia, this is a structurally different problem from the EU or US: it is not primarily about disclosure or risk assessment, but whether a given AI tool may operate against Chinese-held data at all. An unregistered foreign analysis tool cannot lawfully be applied to a Chinese partner’s data set, regardless of how compliant it is elsewhere.

    What does the Council of Europe AI treaty add?

    The Council of Europe’s Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law is the first legally binding international treaty on AI. It opened for signature in September 2024, and early signatories include the European Union, the United States, the United Kingdom, Canada and Japan. It requires ratification by five signatories, including three Council of Europe member states, to enter into force.

    Unlike the EU AI Act, the treaty does not create a detailed compliance regime of its own; it sets human-rights and rule-of-law principles that signatory states must implement through domestic law. For a consortium, it functions less as a rulebook and more as shared vocabulary — a reference point agreements can cite when partners disagree on baseline AI safeguards, even where no national statute yet covers a given use case.

    Where do multi-jurisdiction consortia hit compliance friction?

    The practical friction is rarely about any one regime being stricter than another — it is about the regimes using incompatible triggers. The EU AI Act asks “where is the output used?” US state law asks “which state is the deployer in?” China asks “is this algorithm registered?” The Council of Europe treaty asks “has this state ratified and implemented it?”

    • Data-sharing agreements drafted for one jurisdiction’s risk taxonomy often fail to address another partner’s registration or transparency duties.
    • AI-assisted research tools — plagiarism and integrity screening, generative drafting aids, automated peer-review triage — can simultaneously be “limited risk” in the EU, unregulated in one US state, and require CAC registration in China.
    • Consent and disclosure language for AI use in participant-facing materials rarely satisfies all four regimes’ transparency requirements at once.
    • Governing-law clauses in consortium agreements need to specify which partner’s AI-use obligations apply to shared infrastructure, not just which partner “owns” the data.

    UKRI, Horizon Europe consortia and cOAlition S-aligned funders increasingly expect applicants to describe how AI tools are governed across all partner sites, not only the lead institution’s — making this mapping exercise a funding-eligibility question, not only a legal one.

    Answer-first Q&A

    Are there any global AI regulations?

    No single binding global AI law exists. The Council of Europe’s Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law, opened for signature in September 2024, is the first legally binding international AI treaty, but it needs ratification by five signatories, including three Council of Europe states, before it takes effect.

    Which countries have the most AI regulations?

    The European Union has the most comprehensive statutory AI framework via the EU AI Act, while the United States has the largest volume of measures once state activity is counted. In the 2025 legislative session, all 50 states introduced AI bills and 38 enacted measures, per the National Conference of State Legislatures.

    Which countries have no AI-specific restrictions?

    Several jurisdictions, including the United Arab Emirates and Saudi Arabia, rely on voluntary principles and sector guidance rather than a dedicated AI statute, though both run active national AI strategies and are expected to formalise binding rules as adoption accelerates. Partners based there face fewer AI-specific duties, but other data laws still apply.

    What should multi-country consortia do next?

    No convergence toward a single global AI standard is likely before 2027. The EU AI Act’s high-risk obligations continue phasing in through 2026 and 2027, US preemption litigation remains unresolved, China’s registration regime keeps expanding, and Council of Europe ratifications will accumulate gradually. Consortium agreements that hard-code today’s rules will need scheduled review clauses, not one-off sign-off.

    Research administration teams should treat AI-use disclosure as a standing agenda item in consortium governance, map each partner institution against the table above at project start, and build AI-tool review into existing data-sharing and research administration workflows rather than a separate compliance track.