Artificial intelligence (AI) for academic research at Taylor & Francis

Enhancing the publication process with AI

AI tools help everyone involved in the process of publishing scholarly content – especially authors, editors, and peer reviewers – to work more efficiently. AI lets them focus on tasks where they can add the most value.

Task automation

With human oversight, AI can automate recurring tasks such as manuscript processing, identifying plagiarism, copyediting, proofreading, and indexing.

This eases editor and reviewer workloads, reduces errors, and helps identify fabricated and manipulated data to protect the integrity of research.

AI tools we use include:

• Cactus PaperPal to support language editing
• Journal Suggester to match authors and manuscripts to the most appropriate journals
• CrossRef Similarity Check to check for plagiarism
• Imagetwin to check images for plagiarism, duplication, and manipulation

Identifying reviewers and subject experts

Under human supervision, AI can aid peer review by matching manuscripts with suitable reviewers.

We use Reviewer Locator, a Clarivate tool, to find reviewers. It searches Web of Science data (publications, citations, and reviews) and can recommend up to 30 reviewers from a database of over 28 mn authors.

We also use AI tools to find subject experts when recruiting editorial members and commissioning new authors. Humans will review suggestions and make the final decisions.

Making research easier to discover, access, and understand

AI makes it easier for researchers and readers to find relevant and trustworthy content.

Summarizing content with AI

We are experimenting with using AI to create overviews of complex subjects such as Medicine and Engineering to make them easier to understand and discover via search engines.

Tagging content

We use AI tools to tag content. This makes it easier for readers to find the research they need. It also speeds up the publication process.

Protecting the integrity of research

Although AI has many benefits, it can negatively affect the quality and integrity of research if used without human supervision or by bad actors.

AI-based studies can introduce biases, errors, or inconsistencies in data, methods, and results. They can also manipulate or misrepresent the content or citations of other publications.

As well as this, AI can enable the creation of fake or deceptive content, such as deepfakes, synthetic text, or image forgery. This can mislead readers and harm the reputation of authors and publishers.

Generating or modifying content with AI creates challenges for determining the ownership, attribution, and licensing, especially when research includes content by humans and AI.

In some cases, AI providers may reuse the input or output data from user interactions. This could potentially infringe on the rights of authors and publishers.

How we manage AI risks

Our AI Policy sets out clear expectations for authors, editors, and peer reviewers on how to use and attribute AI in their work.

Our Publishing Ethics and Integrity team provides training programs, guidance, and support. This ensures authors, editors, and institutions have the latest insights, knowledge, and tools to conduct, communicate, and validate research to the highest standards.

Partnerships in the sector with organizations such as the Committee on Publication Ethics (COPE) and the International Association of Scientific, Technical, and Medical Publishers (STM) help make sure we stay on top of how bad actors are using AI to produce fraudulent research.

Partnerships with AI developers

Large language models (LLMs) are AI systems that can process, understand, and generate human language.

Academic knowledge has a fundamental role to play in the development of AI models, including LLMs, that are used both by researchers and the world at large. Carefully considered partnerships with AI developers have the potential to help the discovery of knowledge, impact of research, and effectiveness of learning.

LLMs are trained on massive amounts of data. Academic knowledge is essential for improving the relevance and performance of AI models for academic stakeholders and the wider world. The content LLMs generate is more likely to be inaccurate if the sources they use are not factually correct or scraped from publicly available web pages and documents. We are working with trusted technology partners to improve AI models. This work involves ensuring proper safeguards are applied which include:

• Protecting content from unauthorized access or use
• Limiting the reproduction of verbatim text
• Ensuring authors and other rightsholders receive royalties in accordance with the licensing terms and royalty statement periods in their contracts

Research and resources on AI

We are a leading publisher of AI-related research and content. We have a diverse portfolio of journals, books, and online resources that cover the latest developments, applications, and implications of AI across disciplines.

Journals and books

We publish more than 300 journals that focus on AI or incorporate AI methods in various fields in computer science, engineering, mathematics, medicine, social sciences, and humanities. This includes the world’s first research journal dedicated to Psychology and AI.

We are a leading publisher of book content on AI in computer science, information technology and engineering. We have published over 500 titles on AI and the list is growing rapidly. This includes an AI ethics books series with the Association for the Advancement of Artificial Intelligence (AAAI), the world’s most prestigious AI society.

Policy documents and news articles have mentioned Taylor & Francis books and journal articles about AI more than 65,000 times according to Altmetric.

Online AI resources

We showcase the innovative and interdisciplinary use of AI in research and teaching, via the AI Hub, the AI and Ethics Hub, and the AI and Machine Learning Collection.

Sharing insights and expertise on AI

The Taylor & Francis Content Team curates informative and thought-provoking content on the latest trends and discoveries of AI technology on the Insights blog.