Researcher to Reader 2026 Conference Report

Researcher to Reader Conference Report (24–25 February 2026)

Practical notes for peer review operations and research integrity workflows

Researcher to Reader is built for participation. Speaker introductions are minimal, workshops are central, and breaks offer a choice between networking and optional lightning talks. For editorial office and peer review operations professionals this was an ideal program. Rooted in workflow reality, its talks, panels and workshops covered trust signals and triage, reviewer availability, tools, and author communications – all top of mind for those working in this space.


 

Day 1 – Tuesday 24 February 2026

Keynote: Open and Secure Research (Dr Leslie McIntosh, Digital Science)

Leslie McIntosh’s keynote separated “open” from “trusted”, positioning research integrity, research security, and public trust as interlinked rather than separate agendas. Her framing drew on her FoSci work and the idea of a “battle for knowledge”, where power is exercised through the narratives and legitimacy created by the scholarly record.

A strong operational thread for journal teams was the cognitive reality of the current environment: decision fatigue and signal overload are not incidental; they shape what editors, reviewers, and readers can realistically evaluate. McIntosh used “smart” technology as an analogy for how overcomplication consumes attention and pushes systems toward what is easiest to measure rather than what is most consequential.

For editorial offices, the keynote translated into a practical question: what “trust markers” are being captured, validated, and surfaced consistently? The implied challenge was not simply to collect more verifiable data points, but to build workflows that make correction possible at scale.

 

Workshop B: Using Incentives to Improve Peer Review (Dr Alejandra Clark, The Company of Biologists; Dr David Maslove, Queens University; Dr Tommaso Valletti, Imperial College London)

Workshop B ran as a mini-sprint across the conference: three working meetings (discussion plus small-group design work) ending with a day 2 plenary playback.

The first meeting stress-tested the usual assumptions behind reviewer scarcity: are researchers declining more often or are journals still drawing too narrowly from familiar networks; would widening pools in line with where research is being produced help; and how do editors build confidence in reviewers outside their personal circles when trust feels strained. The discussion kept returning to the same mismatch: heavy reliance on unpaid labour with limited recognition. Paid peer review was debated, with noticeably fewer blanket objections than in similar pre-COVID conversations. Some participants reported their practical experience of paying peer reviewers with positive outcomes.

The second meeting, led by economist Dr Tommaso Valletti, added a blunt warning: incentives can “solve one problem and create another.” Groups interrogated ideas like who pays/benefits in a paid peer review system, who can game it, is it reversible, and what trade-offs are being made between speed, quality, and fairness.

By the third meeting, two proposals were refined for plenary playback: a reciprocity commitment model (authors commit to reviewing within a defined period if their paper is sent for review) and a monetary “marketplace” model (a cross-publisher platform where authenticated reviewers can bid to review manuscripts, with mechanisms to evaluate review quality).

Financial incentives for peer review have moved from taboo to plausible at speed. If paying reviewers is moving from theoretical to operational, then a consistent mechanism by which to assess delivered review quality becomes key.

 

Roundtable: Making Peer Review Better for Researchers (Jayne Marks, Maverick Publishing Consultants (moderator); Dr Ritu Dhand, Springer Nature; Dr Gareth Dyke, Prophy.ai; Dr Chris Leonard, Cactus; Dr Stewart Manley, Maynooth University)

This roundtable asked what authors and readers actually want from peer review, including how to make processes faster and more efficient, whether support should begin earlier in the research lifecycle, how to widen reviewer pools globally (including Asia), and what early career researchers need to trust the record.

Q&A sections around perceived blockers to progress became pointed, as researcher contributors argued that publishers can inhibit peer review innovation through conservative policy and workflow inertia, while others challenged how “quality” judgements can drift into uneven practice across journals and research areas. Accusations of poor editorship and bias surfaced, alongside the recurring gap between public expectations of peer review as a truth test and the reality of peer review as an iterative and variable improvement process.

For operations teams, the practical lesson is familiar but important as ever: opacity and inconsistency are interpreted as bias. Clear criteria, consistent application, and legible decision rationales reduce escalation heat, protect reputation, and maintain community trust.

 

Presentation: Facing AI Futures Together (Professor Keith Webster, Carnegie Mellon University Libraries)

Keith Webster framed AI, at root, as a forecasting tool. He set up two “critical uncertainties” that will shape what scholarly communication becomes next: societal intentionality (how far we regulate, critique training data, correct bias and govern AI, and whose values drive that governance) and societal adaptation (how deeply AI reshapes institutions and day-to-day research practice).

From these axes he mapped four futures. A values-driven, high-adoption world produces Democratised and socially integrated AI: open knowledge flows designed deliberately around shared values, with publishers acting as quality signals — and, he argued, the hardest outcome to realise. A values-driven, low-adoption world produces Consumer-oriented AI: AI behaves as an independent collaborator and human oversight gradually recedes. Ad hoc, market-driven, low-adoption leads to Laissez-faire AI: rapid uptake with weak governance, bias and fragmentation, and a polluted knowledge base. Ad hoc, market-driven, high-adoption produces Autonomous AI: AI becomes collaborator or leader, human oversight retreats further, and the scholarly record fragments, raising the uncomfortable publisher question of who the “reader” even is.

For editorial offices, these scenarios are a prompt to redesign for provenance, consistency, and post-publication accountability, because discovery and “reading” are shifting outside journal control while trust still depends on what the journal can evidence.

 

Debate: Artificial Intelligence Tools “AI Tools Will Provide a Net Benefit to Scholarly Communication” (Rick Anderson, Brigham Young University (moderator); Darrell Gunter, Gunter Media Group (for the motion); Amanda Licastro, Swarthmore College (against the motion))

The audience began optimistic (80% expecting AI to improve scholarly communication) and ended more sceptical (52% expecting it to make things worse), a swing driven largely by arguments about provenance, governance, bias, hidden labour and incentives, and the risks of trading long-term control for short-term convenience.

For peer review operations professionals, we are again reminded that “AI benefit” depends on governance, transparency, and the ability for humans to contest and correct outcomes when tools are wrong.


 

Day 2 – Wednesday 25 February 2026

Panel: Rethinking Peer Review (Dr Jonny Coates, Rippling Ideas (moderator)); Dr Katherine Brown, The Company of Biologists; Thomas Guillemaud, Peer Community In; Dr Peter Rodgers, eLife)

The panel discussed journal-independent, community-led peer review models linked to preprints (including approaches such as eLife, EMBO, Review Commons and Peer Community In), aimed at earlier dissemination of work, reduction of review waste and enabling reviews to travel with a manuscript.

Audience discussion stressed that slow adoption is driven less by publisher resistance and more by author ambivalence: major publishers have funded trials, but take-up has rarely been sufficient to shift default behaviour. Panellists agreed this reflects the difficulty of reaching time-poor researchers – broadcast marketing has limited impact, while genuine engagement is slower and more costly, especially for smaller organisations. “Scooping” concerns were addressed as largely a misconception, since preprints establish priority via DOI; questions from the floor also showed that even basic preprint mechanics are not yet well understood by the research community.

Editorial office takeaway: portability is attractive, but it only works at scale if journals invest in clear author guidance and engagement that changes behaviour, not just promotion.

 

Case Study: Business Models – S2O: The Good, the Bad and the Ugly (Stacey Burke, APS (moderator), Dr Dennis Brown, APS; Professor Wolfgang Kuebler, Charité–Universitätsmedizin; Lou Peck, The International Bunch)

The American Physiological Society (APS) shared their experience with Subscribe to Open (S2O). Their early approach prioritised renewals, with an expectation that value would naturally extend to authors and readers. Feedback revealed confusion, prompting experiments with lighter feedback tools and revised messaging.

Slides and discussion emphasised clearer articulation of organisational purpose (highlighted by APS’s successful “Publish with Purpose” campaign), and the usefulness of the S2O Community of Practice for support and ideas.

For editorial offices, new business model messaging and unfamiliar eligibility rules often land as author queries, increased support load, and additional friction at submission and post-acceptance touchpoints. Get the messaging right and these capacity issues reduce.

 

Presentation: Publication Formats – Beyond the Journal Article (Christian Kaier, Graz University Library; Dr Lisa Schilhan, Graz University Library)

Kaier and Schilhan presented survey findings from 600+ researchers on publication formats beyond the traditional article: data publications, software and code, replication studies, repositories/databases, teaching materials/OER, and science-to-public outputs.

Slides translated the implications into stakeholder actions: libraries supporting infrastructure and cataloguing; funders recognising diverse outputs in assessment; publishers developing models, metadata schemes, and QA processes for novel outputs.

 

Presentations: AI in a Global Context – Does AI Promote Equity or Division? (Nikesh Gosalia, Cactus Communications)

Nikesh Gosalia opened with Kodak’s “Shirley card” story as a clear illustration of how bias becomes a default calibration. He used this to frame a risk for scholarly publishing: AI can lower barriers or scale exclusion, depending on what it is calibrated to “see”.

He anchored the point with an attention-grabbing statistic: 93% of GPT-3’s training data was in English, alongside the observation that tech giants now publish fewer comparable breakdowns. His argument was less about a single model and more about the consequences of opacity: if the sector cannot see what a system is trained on, it cannot predict how it will behave for different researcher communities.

His composite “Dr Amara” workflow illustrated how disadvantage can accumulate across routine steps: discovery gaps for non-English outputs, translation flattening, automated integrity checks that misinterpret English as a second language patterns, and reviewer mismatch leading to lower-visibility publication routes, reinforcing what gets counted as “global knowledge” in future systems.

 

Presentations: AI in a Global Context – Making AI Work for Global Research Integrity (Simone Ragavooloo, Frontiers)

Simone Ragavooloo presented findings from Frontiers’ global survey work and positioned the current challenge as a governance gap rather than a novelty gap. Frontiers’ white paper reports that 53% of peer reviewers already use AI tools, and frames the next phase as requiring trust, training, and governance to keep adoption from becoming unsafe or inequitable.

Two findings strengthened what many in the room were already experiencing operationally: high concern about misuse (71% concerned about misuse by researchers; 45% about misuse by publishers) and low visibility of when and how AI is used in publishing workflows.

Simone described Frontiers’ practical response: creating living guidance for researchers (a playbook-style set of guidelines intended to evolve over time) and piloting a closed-system review assistant to support peer reviewers while also allowing Frontiers to observe how the tool is used in practice and where additional guardrails or training are needed.

The Q&A session following these two talks brought three operational risks into focus: funding and capacity for robust non-English model development remains unresolved, despite early regional initiatives. Human bias persists and can become embedded in routine workflow choices, including patterns of “extra QC” requests clustered by geography or topic. “Human in the loop” can degrade over time: when tools appear consistently reliable, editors and staff check outputs less closely unless workflows deliberately enforce sampling, second-look triggers, and regular audits.


 

Final thoughts

For editorial offices, the recurring themes were:

  • Pair integrity tools with clear, evidence-led author communication and an appeal route.
  • Treat “human in the loop” as something you design carefully and audit regularly.
  • Assume incentives and new output types will force workflow changes.

If these issues are top of mind, The Editorial Hub can be your partner, creating practical policies, templates, and QC routines to support your workflows.

Get in touch for a free, no obligation chat.

 

Research Integrity: Why It Matters and How We Support It at The Editorial Hub

In scholarly publishing, trust isn’t optional – it’s everything. Every paper, every review and every editorial decision contributes to the credibility of the research record. At The Editorial Hub, research integrity isn’t just a policy; it’s embedded in everything we do, from the first submission check to the final decision. Here’s how we help publishers, societies and editors safeguard quality, reduce risks and maintain confidence in every stage of peer review.

 

Why Research Integrity Matters

Research integrity is fundamental to the credibility and reliability of published research and scholarly content. It underpins the trust that publishers, societies and editorial offices place in the work they publish. When integrity is upheld, research can be assessed, built upon and relied on with confidence. When it is compromised, the impact extends across journals, communities and reputations.

For editorial offices, research integrity is not an abstract principle. It is embedded in everyday editorial work, including managing submissions, supporting peer review, applying ethical and editorial policies, and maintaining accurate records of editorial decisions. As submission volumes grow and editorial workflows become more complex, consistency and transparency across these processes are increasingly important.

Publishers and societies are responsible for defining editorial and ethical policies, often informed by recognised industry guidance such as that provided by COPE. The Editorial Hub supports publishers by implementing these policies within editorial workflows, helping to ensure they are applied consistently and in line with client expectations. This support plays an important and vital role in identifying potential concerns early and ensuring they are handled appropriately.

 

The Editorial Hub as a First Line of Defence

We see ourselves as a critical first line of defence for research integrity. Our team combines deep expertise in manuscript assessment, preliminary checks and peer-review management to identify risks such as papermills, peer-review manipulation, plagiarism, image/data issues and ethical non-compliance early in the process. By surfacing concerns before peer review begins, we allow reviewers to focus on scientific merit, improving both efficiency and reliability of the editorial process while reducing the burden on reviewers.

At The Editorial Hub, research integrity is reinforced through close alignment with publisher and society policies and through well-defined editorial processes. Our work is shaped by our clients’ requirements, ensuring that expectations around ethics, authorship, conflicts of interest and transparency are reflected accurately throughout the editorial processes we undertake. Applying these standards consistently is essential to maintaining confidence in editorial outcomes.

 

Research Integrity Leadership

Our Research Integrity Manager (Jenny Bright) leads our work to uphold the highest ethical standards in scholarly publishing. This role includes developing processes and assessing tools to detect issues such as plagiarism, image manipulation, papermills and peer-review manipulation, as well as providing guidance and resources to our team, editors and clients on best practice. Jenny represents us in wider industry discussions, sharing insights and contributing to evolving best practices that maintain trust and credibility in peer review.

Clear collaboration and communication are central to this approach. Our editorial offices work closely with journal editors and publisher contacts to support informed decision-making. Maintaining clear and open lines of communication, accurate documentation and shared understanding of responsibilities helps ensure that issues are addressed efficiently and in accordance with agreed policies.

 

Supporting Editors and Reviewers

Peer review remains a key component for maintaining quality and credibility. Through structured editorial support, our team manages peer review workflows that are fair, transparent and well documented. This supports editors in coordinating reviews, recording decisions and maintaining clear oversight of the review process. We also provide structured reviewer support, including templates and integrity checklists, helping reviewers systematically identify potential issues and improve the consistency and quality of peer-review reports.

 

Tools, Automation and AI Oversight

The increasing use of artificial intelligence across research and publishing introduces additional considerations for research integrity. Expectations around the use of AI tools in submissions, peer review and editorial processes continue to evolve. We actively review the landscape of AI and research integrity tools, trial and test emerging technologies, and advise clients on those best suited to their journals and workflows. By combining automation with human oversight, these tools allow us to detect potential concerns early and reduce reviewer burden while maintaining robust editorial judgement.

 

Maintaining the Research Record

Maintaining the integrity of the research record is an ongoing responsibility. It requires sustained collaboration between publishers, societies and editorial support providers, as well as the flexibility to adapt processes as standards, technologies and expectations develop.

At The Editorial Hub, we see research integrity as a shared commitment and collaboration with the clients we support. By working in line with publisher policies, communicating effectively with editorial stakeholders and applying consistent editorial practices, we help publishers and societies proudly maintain confidence in their editorial processes and the content they publish.

 

Shaping Research Integrity Standards

Through our active industry engagement, including membership in COPE, ALPSP and STM, as well as participation in seminars and events, we contribute to shaping best practice in research integrity and ensure our team remains at the forefront of evolving standards.

 

Hannah Gardiner, Managing Editor / Social Media Manager

 

Research integrity after STM 2025: what changes on the editorial desk?

STM’s Innovation & Integrity Days (9-10 Dec 2025) in London were framed as a multi-stakeholder conversation, but a lot of what was discussed lands directly on editorial offices: managing editors, journal coordinators, assistant editors and publisher-side integrity staff. This isn’t just about high-level policy, it’s about how submissions are screened, how editors are supported, and how conversations with authors and reviewers are going to change.

Here’s what the two days signal for people working on the ground.


  1. Integrity tools are becoming part of core workflow, not an optional add-on

The Innovator Fair made one thing very clear: the tools are converging on the editorial desk.

Systems like the STM Integrity Hub, DataSeer, Smart Publish, PubShield, Enago Reports, ReviewerZero and others are all pushing toward the same pattern: one embedded view of multiple integrity signals, rather than five different logins and exports. For editorial offices this means:

  • More signals on each manuscript (images, text, references, data, identity, networks).
  • Fewer standalone dashboards, more integration into systems like Editorial Manager, ScholarOne and the like.
  • A shift from “should we run this check?” to “how do we interpret this combined report?”

On the desk, this changes the job from doing checks to triaging them. You’ll need:

  • Clear SOPs for what to do with specific patterns of flags.
  • Thresholds for when to escalate to research-integrity specialists.
  • Templates for neutral, factual queries to authors about technical findings.

The technology is the start; the real work is designing consistent responses around it.


  1. Checks will move earlier and go deeper

Several sessions – on the Integrity Hub, image/data authenticity, and the sleuths’ work – pointed in the same direction: pre-publication is where the real defence has to happen.

For editorial teams this likely means:

  • More pre-submission or pre-review screening (plagiarism, duplication, image anomalies, strange references, “AI nonsense” language).
  • More frequent requests for original images and underlying data, not just prettier figures.
  • Stronger expectations that journals know what they will and will not accept as “provenance”.

You should expect author guidance to tighten. Instructions for authors will need to be clearer about:

  • What editors and offices will ask for in problematic cases
  • How images and data should be stored so that provenance can be shown
  • What happens if authors cannot or will not supply originals

For support staff, the practical impact is more back-and-forth before peer review even starts, with more technical documentation attached to each case.


  1. Identity checks are coming – and may help more than they hurt

The digital identity discussions were blunt: publishing runs millions of articles a year with almost no “know your customer” equivalent.

Static affiliation is no longer enough. Shell institutions, exaggerated affiliations and identity spoofing are all in play. At the same time, there are serious experts working outside traditional academia who struggle to enter editorial workflows because they lack a neat university label.

For editorial teams, the likely direction is:

  • More structured identity verification at certain points (e.g. new corresponding authors, guest editors, reviewers from unknown institutions).
  • More use of ORCID, institutional attestations and simple ID checks to reduce spoofing.
  • Possible TSA pre-check style “lanes” where higher-assurance identities enjoy fewer manual checks.

There was also a useful reality check: in daily life, people already accept identity checks everywhere – banking, smartphones, even renting an e-scooter. Relative to that landscape, asking for basic ID assurance around work that influences patient care, public policy and now AI training data is hardly extreme.

Your role will likely include:

  • Explaining calmly why extra checks are being introduced.
  • Ensuring any new ID workflows are documented and applied consistently.
  • Watching for equity issues – for example, how new requirements affect authors without strong institutional backing.

  1. Provenance isn’t abstract – it will change how you handle images, data and supplements

The “authenticity of images and data” sessions were clear: we are not going to win an arms race against future image and data fakery by detection alone. As AI-generated content improves, some manipulation will be practically undetectable.

Editorially, this shifts focus from “spot the fake” to “prove the workflow”. That means:

  • Treating original images, lab metadata and data files as normal parts of the submission ecosystem, not rare exceptions.
  • Taking supplementary materials more seriously: they are where many critical data now live, and where LLMs will mine heavily.
  • Being ready to ask, “Where did this image/data come from? Can we see the unedited version? Do we have enough to be confident?”.

Several ideas that may reach the desk:

  • Journals asking for signed or registered originals in higher-risk fields.
  • Data-focused reviewers (perhaps postdocs?) involved alongside traditional peer review.
  • Badges or indicators in articles where provenance checks have gone beyond the minimum.

For staff, this means more file handling, more liaison with production and more judgement calls about when “enough provenance” has been demonstrated.


  1. Working with institutions and sleuths will need structure, not ad hoc goodwill

Panels on publisher–institution collaboration and the “session with the sleuths” exposed real friction:

  • Institutions see publishers as slow or opaque.
  • Publishers see institutions as unresponsive or defensive.
  • Sleuths see both as under-resourced and often silent after detailed reports.

For people inside editorial workflows, there are three practical implications:

  1. Case handling needs clearer channels
    • Know who handles serious integrity cases in your organisation.
    • Know how and when to contact institutions (and what not to send).
    • Keep a clean audit trail: dates, decisions, communications.
  2. Sleuth reports need visible triage
    • Acknowledge receipt promptly.
    • Log reports in a trackable system.
    • Close the loop where possible (even if the outcome is “no action, here’s why”).
  3. Backlog management will move from “nice to have” to expectation
    • Tools that find articles citing retracted work are here.
    • Editorial offices may be asked to support back-catalogue sweeps, not just new submissions.
    • You’ll need criteria for what gets investigated first.

The message from sleuths was straight-up: if publishers want their help, they need to show that it leads somewhere.


  1. Policy frameworks are under strain – you will feel that at case level

Guideline bodies and cross-publisher initiatives are updating retraction guidance, expressions-of-concern policies and support for editors under pressure. At the same time, we are seeing on the ground that the volume and sophistication of fraud is testing procedures designed for a now bygone era.

On the ground, this shows up as:

  • Long, complex casework in resource-hungry conditions.
  • Risk/reward of following industry best practice called into question.
  • Grey areas where existing policies say little (paper mill adverts, recycled images).
  • Pressure from multiple directions: authors, institutions, sleuths, media, internal leadership.

The emerging direction may well be:

  • Keep due process and fairness, but
  • Develop faster tracks for situations with clear, well-documented evidence.
  • Be explicit about which pathway a case is on, and why.

Editorial staff may not write the policies, but you will enforce them, document them and explain them repeatedly. Expect more attention to internal training on this.


  1. What skills and behaviours will matter most?

Across both days, a pattern emerged of what the system increasingly needs from editorial teams:

  • Triage, not just task execution
    Ability to interpret multi-signal reports and make sensible calls about escalation.
  • Calm, precise communication
    Neutral, factual language in difficult emails to authors, reviewers and institutions.
  • Comfort with tools and data
    Not coding, but understanding what image, text and data checks can and cannot tell you.
  • Process literacy
    Knowing your own policies, where they’re strong, and where they’re silent – and documenting decisions accordingly.
  • Resilience and boundaries
    Integrity work carries emotional load: accusations, counter-accusations, legal threats. You’ll need support and realistic workloads, not heroics.

The upside: the work is becoming more visible and more valued. The downside: it is also becoming more complex.


In summary

The STM Innovation & Integrity Days didn’t just showcase new tools or restate familiar principles. They described a future where research integrity is baked into every stage of the editorial workflow: who you accept manuscripts from, what you ask for, how you route signals, and how you interact with institutions, funders and sleuths.

For editorial professionals, this is not a bolt-on extra; it is the next evolution of the role. Calls from the stage for substantially increased investment in research integrity teams were met with clear, audible approval from a senior audience. That’s a signal: organisations know they need more capacity at the coal face. For anyone looking to deepen or reorient their career around integrity, there is likely to be more demand, more responsibility and, over time, more formal roles in this space.

Publication Integrity Week 2025: Key Themes and Reflections

Our Research Integrity Manager, Jenny Bright, was an active member of COPE’s programme of discussions during Publication Integrity Week 2025. The week brought together global voices to explore the evolving challenges and opportunities in scholarly publishing, with each themed day highlighting the intersection of ethics, technology, inclusivity, collaboration, and innovation.

Jenny distilled the week’s discussions into several key insights:

  • The evolving landscape of publication ethics. The week began with a reminder of ‘Why ethics matter in scholarly communications’. COPE’s early leaders discussed how challenges have grown since COPE was founded in 1997, but with a growing community of over 14,000 members, there is a building strength to ensure continuity of COPE’s mission. The discussion on ‘Commercialisation of research integrity’ highlighted how profit-driven misuse and policy gaps can undermine trust in the system. The session on ‘Technology for transparency’ promoted discussion on FAIR data and open research.
  • Diversity, equity, inclusivity, and accessibility in practice (DEIA). Discussions focused on how integrity is inseparable from inclusivity. Case studies showed how journals are adopting policies to ensure equitable representation in authorship, peer review, and editorial boards. Integrity is not just about avoiding misconduct, it’s about proactively creating fairer systems.
  • Collaboration and community building. Collaboration was a key theme throughout the week, and it was interesting to hear about both cross-institutional and cross-national initiatives. Panels explored how shared standards and joint investigations can strengthen integrity globally. The message was clear: no single journal or institution can tackle misconduct alone.
  • Upholding publication integrity; the role of universities and research institutes. Sessions on institutional responsibility highlighted the need for trust and integrity, not just compliance during research. Institutions are developing frameworks and training for students, setting the tone for ethical publication of scientific research.
  • Innovations, solutions, and the road ahead. The final day explored new publishing models and the tools available to editors and publishers to combat misconduct. It highlighted the evolution of AI as both a tool and a test for ethical frameworks and the need for clear cross-publisher policy.

The week was an insightful journey, from the roots of publication ethics, through present challenges, to future possibilities; giving a renewed conviction that integrity is a shared responsibility across the academic publishing community.

Rethinking Peer Review in the AI Era (Peer Review Week 2025)

Jenny Bright, our Research Integrity Manager, was fortunate to attend a number of webinars and talks during Peer Review Week 2025. Here, she shares her reflections and key takeaways from the discussions.


AI in Peer Review: Challenge or Opportunity?

The key talking point across the week was artificial intelligence (AI), with a primary focus on “Rethinking Peer Review in the AI Era.”

We discussed how AI is reshaping the peer-review process – and while it presents a number of challenges, it’s important not to overlook the opportunities it brings.

The consensus was clear: AI is here to stay. Publishers must tackle it head-on, implementing robust tools and procedures to identify AI use, ensuring publication integrity is maintained and that AI involvement is made transparent to both authors and readers.

Where AI Is Making an Impact

AI is now prevalent in academic publishing in many areas – from AI-generated articles and manipulated images or data, to the use of AI for conducting reviews, and the tools publishers use to detect and combat AI misuse.

As AI tools become increasingly embedded in academic workflows, questions have been raised around ethics, bias and transparency, such as:

  • How can AI enhance reviewer efficiency without compromising integrity?

  • What ethical frameworks should guide AI-assisted peer review?

  • Which aspects of peer review must remain distinctly human?

My Key Takeaways from Peer Review Week

  • AI in peer review should be a support tool, not a replacement.
    Experts agreed that while AI can streamline tasks like plagiarism detection, citation checking, and reviewer matching; human judgment and critical analysis remain essential for evaluating novelty and scientific rigour. It was agreed that AI and humans together can produce an effective peer-review process.

  • Industry-wide standards are necessary.
    Every publisher is currently handling AI differently, without an overarching set of guidelines. There’s a growing need for clear, cross-publisher policies on how AI should be managed and used throughout the peer-review process, with regular monitoring and review as AI evolves.

  • Combatting papermills and identifying data/image manipulation.
    AI is a positive and invaluable tool to help detect patterns of misconduct, such as AI use, papermill submissions and citation stacking. There are some pioneering tools in development, and used correctly they can identify ethics concerns in a shorter timescale than would be humanly possible, however they must be tested and used responsibly to avoid false positives.

  • Reviewer training and support.
    Institutions must educate reviewers on AI tools and their limitations to ensure informed and ethical use.

  • Preserving trust and transparency.
    Across the week, the publishing community emphasised that trust in peer review is fundamentally important, and there must be openness about how decisions are made – especially when AI is involved.

In Summary

The emergence of AI in peer review, and academic publishing more broadly, is both a threat and an opportunity.

The peer review landscape is changing (and will continue to change) as a result of AI development. We must identify the risks, establish robust safeguards at every potential point of AI integration, and harness AI’s ability to detect anomalies and patterns to combat misconduct.

By doing so, we can ensure research and publications remain high-quality, transparent and trustworthy across scholarly publishing.

Raising the Quality and Reproducibility of Data-for-Policy Research

In today’s fast-moving research landscape, maintaining high standards of transparency, reproducibility and ethical practice is more important than ever. We’re pleased to share a new video featuring our Research Integrity Manager, Jenny Bright, in which she discusses key challenges, practical steps and the cultural shifts needed to keep research robust and trustworthy.

The latest Data for Policy Fireside Chat explored how to make data-for-policy research more transparent, reliable and reproducible.

In the session, titled “Raising the Quality and Reproducibility of Data-for-Policy Research,” experts – including Jenny Bright (Research Integrity Manager, The Editorial Hub) – discussed practical ways to strengthen research integrity in data-driven policymaking and academic publishing. The conversation focused on the Editorial Reference Handbook, developed through the Tier 2 Project, which promotes FAIR (Findable, Accessible, Interoperable and Reusable) data practices. Speakers shared lessons from Data & Policy’s pilot, including the use of reproducibility checklists and clearer data-availability statements.

Key outcomes show real progress:

  • Every Data & Policy article now includes a detailed data-availability statement.

  • Dataset links with DOIs rose from 18% to 30% during the pilot.

  • Authors are increasingly transparent about restricted data access.

  • Collaboration between journals, publishers and data curators is embedding reproducibility into research publishing.

Together, these steps are building greater trust and credibility across the data-for-policy field.

Watch the full conversation below:

https://youtu.be/rW58p7Enhg0?si=TLOw4YyPmTcLP1ov