Liangping Ding

Artificial Intelligence (AI) is rapidly changing how scientific research is conducted, offering tools that can assist with tasks such as literature reviews, data analysis, and even writing. However, we still know very little about how the adoption of AI is actually shaping the behaviours, outputs, and career paths of researchers. This project aims to fill that gap by investigating how AI has transformed academic knowledge production. This research focuses on the biomedical sciences, a field of critical importance for public health and innovation, and will explore how AI tools are being adopted by scientists at different stages of their careers, as well as the impact of AI adoption on scientists’ productivity, creativity, research diversity and career trajectories. For example, it will look at whether junior researchers—who may be more digitally fluent—are quicker to embrace AI than their senior counterparts, and whether this leads to more equitable research outcomes or widens existing gaps. The study will pay particular attention to differences across gender, career stage and institutional affiliation, to understand whether AI acts as a leveller or deepens inequalities in academia. To do this, the project will draw on large-scale publication data (using sources such as PubMed knowledge graph and Open Alex). Additionally, a specific case study comparing PhD graduates in the UK and Japan based on large scale dissertation data will offer international insights into how AI is affecting doctoral training. The aims of the project are threefold. First, it will advance the growing field of metascience by providing robust evidence on how AI affects research productivity, creativity, diversity, and the career trajectories of scientists. This will involve using state-of-the-art AI tools—particularly large language models and Generative AI—not just as objects of study, but as methodological instruments to analyse large-scale scientific datasets. By using AI to study AI, the project will explore the capabilities and limitations of these models in extracting meaningful patterns from scientific texts, enabling novel forms of analysis that go beyond conventional bibliometric techniques. Second, the project will develop new conceptual and analytical frameworks to understand how AI influences the organisation of scientific work—shaping collaboration networks, the adoption of new research methods, and the evolution of research agendas. Third, it will generate actionable insights to inform the design of more inclusive, human-centred AI tools. By identifying adoption patterns, and demographic disparities, the project will assess whether current AI systems adequately support diverse researchers and research settings. The findings will be relevant to universities, research funders, and policymakers who are seeking to understand and guide the responsible integration of AI into science. They will be shared through high-profile academic publications and conferences and translated into practical recommendations for research institutions and policymakers. Insights from this project will be disseminated to relevant stakeholders, including the Greater Manchester Combined Authority (GMCA), DSIT government office for science (Go-science), to ensure the results contribute to digital inclusion, talent development, and responsible innovation. Ultimately, this research will improve our understanding of how AI is reshaping the nature of scientific work. It will offer new tools and evidence to support more equitable and effective use of AI in research, helping to ensure that its benefits are widely shared and that the next generation of researchers is equipped to thrive in an AI-driven world.