The TRAIN Act grants rightsholders the ability to petition courts to subpoena developers to release generative AI training data. The CLEAR Act would require generative AI developers to disclose, available via a Copyright Office database, detailed summaries of the copyrighted works used during training for all publicly accessible models.

STM CEO Caroline Sutton provided supportive quotes regarding the goals of both of these bills, and will be working with the bills’ sponsors on details if they move forward in the legislative process.

The post STM supports transparency in AI training appeared first on STM Association.

To address this issue, STM has released a report “Recommendations for a Classification of AI Use in Academic Manuscript Preparation” in September 2025. 

Today, we’re announcing an exciting follow-up activity that addresses this issue from a broader perspective through collaboration with key partners in the academic enterprise.

To support a shared understanding of how AI should be disclosed in research, STM is part of a joint harmonisation initiative to work towards a Global Reporting Standard for AI Disclosure in Research, together with the Committee on Publication Ethics (COPE), the International Science Council (ISC), and the Global Young Academy (GYA) as key partners. This initiative will form the focus track of the World Conference on Research Integrity (3–6 May 2026, Vancouver). 

To ensure that the development of this reporting guideline reflects a broad range of perspectives, STM and the other key partners invite their members to contribute through three consultation rounds — the first round is open now through 28 February 2026: Mapping the needs for AI disclosure, yielding a preferred format.  

To that end, we invite STM members and community to participate in the first consultation round, as it is vital that the format of this standard matches current experiences and expectations of the publishing industry. On behalf of the organising partners, we invite you — as an STM Member Organisation — to participate by taking two actions: 1/ review some preparatory reading and discuss it with colleagues (in the editorial office or beyond) and then 2/ submit a summary of your collective reflections via the form below. 

Access the preparatory questions and reading 

Submit your input here by 28 Feb, 2026

All questions and remarks you may have, can be directed to Bert Seghers via email and/or Hylke Koers via email.  

We appreciate your commitment to this important effort to build a shared, global understanding of AI disclosure in research.

The post Global reporting standard for AI disclosure in research: first consultation is open appeared first on STM Association.

In an article on growing threats to research integrity, Times Higher Education covers STM’s report Safeguarding Scholarly Communication: Publisher Practices to Uphold Research Integrity. The article describes how publishers are increasingly focused on identifying integrity issues before publication—responding to paper mills, AI-enabled fabrication, and coordinated fraud networks—while scaling up research integrity teams and collaborating on shared screening approaches.

THE highlights the STM Integrity Hub as an example of collaboration designed to improve detection capacity across publishers, while noting the continued opportunity to extend benefits to smaller publishers and those operating outside major publishing centers.

Read the full article at Times Higher Education → (subscription required)

Read STM’s report→ available here

The post In the media | Times Higher Education — “Unseen efforts to catch paper mill outputs bear fruit” appeared first on STM Association.

The case centres on Westlaw’s “headnotes” — concise editorial summaries created by legal experts to help users navigate case law. ROSS Intelligence, a now-defunct startup, used thousands of these headnotes to train a competing AI-driven legal research tool, without permission. A lower court ruled that the headnotes are protected by copyright and that ROSS’s copying did not qualify as fair use. ROSS is appealing the decision. 

The Copyright Alliance brief, which incorporates input from STM, argues that Westlaw’s headnotes meet the originality threshold for copyright (not unlike scholarly abstracts, summaries or editorial metadata crafted by STM publishers). It also emphasizes that allowing unlicensed reuse of this kind could undermine the legitimate licensing market for high-quality, curated content — especially when used to train generative AI tools. 

STM supports the brief’s position that copyright protects not just personality-laden works, but also the intellectual expressive works that involve crafting structured, discoverable, and context-rich editorial material. As in STM members’ own publishing workflows, such content is created with skill, creativity, and judgment — and should not be freely copied or repurposed without regard for copyright protections or the value of the underlying work. 

STM joins the Copyright Alliance and other amici in encouraging the court to uphold the principle that copyright supports innovation — including in the development of responsible, licensed AI. 

Delve deeper
 

The post STM supports Copyright Alliance brief in key U.S. copyright case appeared first on STM Association.

Now in its third year (building on the legacy of STM Week), this year’s Innovation & Integrity Days reflected a noticeable shift: more dialogue across traditional boundaries, more urgency around shared challenges, and a stronger sense of common purpose in shaping the future of research integrity and innovation — priorities that reflect this year’s theme “Building Tomorrow’s Research Integrity Framework”.

For those unable to join us in London, here are some of the key takeaways and highlights:

Day 1 Recap

Day 2 Recap

Day One: the STM Innovator Fair

The first day of STM’s Innovation & Integrity Days focused on practical progress in innovation — with a keynote, lightning talks, startup pitches and a panel that explored how technology, infrastructure and collaboration are reshaping research integrity.

In the opening keynote, Professor Rachael Gooberman-Hill (UK Committee on Research Integrity; University of Bristol) set the tone by drawing on her extensive experience to frame research integrity not as a checklist, but as the long-term work of building and maintaining trust — informed by context, relationships, and values.

The lightning talks that followed showcased a wave of innovation already underway — not just isolated tools, but signals of how the ecosystem is adapting. Key themes included:

• Workflow integration: Platforms like KGL’s Smart Publish, Proofig’s PubShield, Enago Reports, ReviewerZero and Hum’s Alchemist Review demonstrated how multiple integrity signals can be brought together in a single editorial interface. DataSeer’s SnapShot tool demonstrated how agentic LLM technology can be leveraged to support the editorial process and Scitility, a Vesalius Innovation Award finalist, added to the picture with article-level risk scoring based on co-authorship patterns and retractions.
• Upstream provenance: Solutions like Veridat’s Bench and VeriMe are moving trust signals earlier in the process — capturing the when, where and how of data creation, and validating researcher identity in a privacy-preserving way. This was echoed in Thesify AI’s pitch — using structured feedback and journal matching to support quality at submission.
• Reviewer capacity and quality: ReviewerOne, Global Campus and SinoScholar (MPS) explored ways to strengthen reviewer pools through better matching, recognition and AI-assisted support. Profectus Academy, the Vesalius Award winner, connects here too — using real-world examples to build reviewer-like feedback into the research grant writing process.
• AI-readiness, risk awareness and trust: From Cashmere’s work on preparing trusted content for LLMs to Otto-SR’s systematic review support and Dandelion’s visual literature mapping, the ecosystem is thinking ahead. AuthentiSci, another Vesalius finalist, approached this from the public trust angle — enabling ORCID-verified researchers to rate and contextualise science reporting.

The innovations on display reflected a shared aim: to embed trust earlier, deeper and more systematically into the research process — supporting not just editors and research integrity staff, but researchers, reviewers and the public.

The day also featured Karger’s Vesalius Innovation Award, spotlighting early‑stage innovation focused on trust and researcher support. Five finalists pitched from more than 100 applications worldwide; you can learn more about the winner Profectus Academy and the other finalists here.

A five-year retrospective on GetFTR offered a compelling testimony of cross-industry collaboration delivering results. Originally launched to streamline access to the version of record from discovery services, GetFTR has evolved into critical infrastructure supporting the discoverability and use of trusted scholarly content in a multitude of ways.

The Fair also created space for practical exchange. Between sessions, attendees connected with exhibitors, explored prototypes and surfaced shared challenges. In parallel, in-person Research Integrity Workshops provided case-based learning in a confidential setting — not covered in this report, but an important part of the day’s impact.

The final session, From Pitch to Partnership, explored how stakeholders assess new ventures. Investors, infrastructure leaders and acquirers shared frank insights: what works, what doesn’t, and how mission-aligned startups can grow without compromising their values.

Taken together, the Innovator Fair pointed to a clear direction of travel: innovation that embeds trust throughout the research lifecycle — supported by smart tools, shared infrastructure, and steady collaboration.

Day Two: Research Integrity Day

Held under the Chatham House Rule, Day Two of the programme was framed as a multi-stakeholder dialogue — to surface shared concerns, test one another’s assumptions, and discuss practical pathways forward. The day kicked off with an update on the STM Integrity Hub, highlighting the progress that has been made in 2025, and then featured a series of panel discussions that explored various aspects of the “Building Tomorrow’s Research Integrity Framework” theme. 

Bringing together publishers, researchers, funders, institutional integrity offices, infrastructure providers and volunteer sleuths, the day focused on how to strengthen trust in research amid evolving threats and expectations. Discussions focused on research integrity, governance and operational practice, and did not cover competitively sensitive topics such as commercial terms. Three key themes emerged:

1/ Shared Responsibility, Misaligned Incentives
Participants acknowledged that every part of the research ecosystem plays a role in upholding integrity — but also faces different incentives and pressures. From grant cycles and publication demands to institutional risk management and business models, aligning integrity incentives emerged as a long-term priority. The discussion pointed to the need for collaboration, not just isolated efforts.

2/ Shift Left: Trust Starts Earlier
There was strong support for moving integrity checks earlier in the research lifecycle. By the time a publisher’s integrity team gets involved, behaviours are already entrenched and corrections are very cumbersome. Funders and institutional representatives described very low numbers of formal complaints relative to portfolio size, with only a fraction upheld, and pointed to weak whistleblower protections, fear of reprisals and unclear reporting routes as key barriers. Suggested directions included treating role-modelling and local culture as central (not just formal training), exploring continuing professional development requirements for integrity awareness, and tightening channels and protections so that raising concerns is not perceived as a risk to one’s career. Funders described extending grant periods, building in generous leave extensions and placing greater weight on environment and process rather than on publication counts alone. Participants, including sleuths, also called for better protections for those who raise concerns, and for local research cultures that treat integrity as a shared value embedded in everyday practice, rather than a box-ticking exercise.

3/ Upstream Provenance and Workflow Integrity
Several sessions carried forward last year’s Innovation & Integrity days theme of going “beyond the manuscript”. New safeguards, in addition to checking manuscripts at the point of submission for indications of integrity concerns, are required to establish the veracity and authenticity of research outputs, especially in the age of (Gen)AI. Two options were discussed specifically: strengthening researcher identity verification, and establishing the authenticity of images and data through new technologies. Participants agreed that it will be increasingly challenging for detection to keep pace with manipulation as tools evolve, which requires us to focus our attention on provenance — capturing how and where data and images are created, and embedding that information into research workflows from the outset. Ideas included trusted research environments, data-focused peer review, and stronger signalling of provenance quality, aligned with Open Science principles. In other words, a shift from publishers investigating if something is false or fabricated, to researchers demonstrating that something is real and genuine through trustworthy workflows – requiring collaboration and new workflows throughout the research ecosystem.

A Way Forward
One clear takeaway from both days: technology can support integrity, as proven by the success of the STM Integrity Hub and the many innovators presenting their work, but it cannot guarantee it. Lasting trust depends on governance, incentives, new technologies, workflows, and especially collaboration.

STM remains committed to supporting that progress — convening dialogue, connecting expertise, and championing efforts that protect and enhance the integrity of the scholarly record.

Special thanks to Lynsey Haire, Academic Publishing Operations Consultant, for her expert summary and contribution to this post following STM’s I&I Days in London.

—

Want to be sure you don’t miss next year’s event? (Hint: it will be in early December 2026 at the BMA House again!) Sign up for our newsletter so you can be the first to know when registration opens — and when virtual learning opportunities emerge. 

The post A recap: STM Integrity & Innovation Days 2025 appeared first on STM Association.

In early November, STM CEO Caroline Sutton spent several days in Tokyo meeting with funders, government leaders, research agencies, and publishing groups — alongside delegates from STM’s Japan Chapter. As in last year’s visit, the conversations were productive, wide-ranging, and grounded in strong local partnerships. And while open science dominated the agenda in 2024, this year the spotlight had clearly shifted to AI, research integrity, and the evolving policy landscape in Japan. 

Government: A New Five-Year Plan and a Changing Landscape 

One of the most important conversations took place with Professor Kohei Miyazono, the new Chief Executive Member of Japan’s Council for Science, Technology and Innovation. With his background as a journal editor and society president, he brings a nuanced understanding of scholarly publishing to the table. 

Japan’s next Five-Year Plan is expected in early 2026 and may include efforts to reverse the country’s relative decline in research output — potentially with an increase in national research investment. 

Discussions also touched on rising political tensions with China, including a growing concern: inter-governmental research co-funded by China may require that 50% of resulting publications appear in Chinese domestic journals. 

Open science, which dominated last year’s visit, received only brief attention — a sign that, with the national policy now launched, ministries have turned their focus to newer priorities. 

Funders: Open Access Progress & AI Guidance 

Meetings with AMED, JSPS, and JST remained warm and constructive. 

• AMED sees open access expanding gradually under the national policy and is now considering how to guide researchers on use of AI — both in research practice and in grant evaluations.

• JSPS anticipates possible increases to Kakenhi grant funding and has recently issued guidance on AI use.

• JST is introducing GrantsData, a new dataset repository that joins a growing ecosystem: J-STAGE, Jxiv, ResearchMap, and JBM. JST also reported increasing requests from LLM developers seeking access to J-STAGE content for training.

The annual JST/STM Seminar drew its largest audience yet (217 attendees), with a strong slate of talks focused on AI’s impact on publishing, research workflows, and data integrity. 

Copyright & Rights Management: New Conversations and a Broader Table 

This year’s agenda expanded to include both the Agency for Cultural Affairs (ACA) and the Japan Book Publishers Association (JBPA). 

ACA reaffirmed that any AI use of copyrighted works that risks harming commercial rights requires licensing under Japanese law. They are now convening creative industries and rights-holder groups across three working groups to develop a Code of Conduct for AI and copyright — an area where STM has offered continued support. 

Discussions also included updates on: 

• SATRAS, which collects and distributes compensation for educational copying (roughly ¥4.8 billion annually), and

• SARLIB, the new system overseeing library transmissions under Article 31 of the Copyright Act, where publishers may opt their titles out of transmission.

Open Access & National Policy Momentum 

With the April 2025 launch of Japan’s national open science policy, agencies signaled they are now turning attention to other pressing issues such as AI governance and research integrity. 

At the same time, STM data presented during the visit showed an interesting trend: the share of subscription articles is increasing while open access output is tapering, despite a growing number of read-and-publish agreements. 

Factors contributing to this include: 

• Institutions asking researchers to pay additional fees even under existing agreements 

• Restrictions at some universities on publishing in non-Q1/Q2 journals 

• Policy flexibility allowing non-immediate OA when a reason is provided

As a result, many articles may continue to appear in subscription journals for the foreseeable future. 

Meanwhile, NII continues its work on metadata integration and infrastructure development through CiNII Research, connecting institutional repositories across Japan. 

Local Publishing Community & Outreach 

Caroline and the STM delegation also met with JBPA leadership for the first time in over a decade — a meaningful step toward deeper collaboration on copyright, AI, and shared challenges facing publishers globally. 

An STM Japan Chapter meeting was held at Wiley’s offices, and Caroline also gave an interview to Science News, reaching 40,000 readers across Japan’s research community. 

In Summary 

Caroline’s 2025 visit to Japan underscored a clear shift in national and institutional priorities: 

• AI is now the central topic of concern across government, funders, and publishers.

• Open science policy implementation is underway, but attention is moving toward research integrity, infrastructure, and practical impact.

• Copyright and rights management are entering a new phase, with Japan taking proactive steps to clarify how AI and copyrighted content can coexist.

• Engagement with Japan’s publishing ecosystem remains strong, with new bridges being built between STM and national associations.

Overall, this year’s meetings reinforced Japan’s importance as a thoughtful, collaborative partner in shaping the global future of scholarly communication. 

The post Inside STM’s November visit to Japan: key themes, takeaways & what’s next appeared first on STM Association.

Selected from a record-breaking number of submissions this year, these innovators showcase some of the most promising technologies and ideas shaping the future of trusted research.

Attendees will experience these innovations first-hand—through lightning talks, live demos on the Fair floor, and also experience the Karger Publishers 2025 Vesalius Innovation Award ceremony, spotlighting five outstanding finalists.

Learn more about this year’s theme and selected presenters — and register to join us! >>

The post Announcing the selected presenters for the upcoming STM Innovator Fair appeared first on STM Association.

The post STM responds to AI labelling consultation appeared first on STM Association.

The post EU adopts strategy for AI in science appeared first on STM Association.

The post White House seeks input on AI regulation appeared first on STM Association.