Why GenAI alone won't drive innovation

Companies are increasingly embedding generative artificial intelligence (GenAI) into their innovation processes, often assuming that stronger AI capabilities will naturally lead to better innovation outcomes. However, this assumption is too simplistic.

Research published in Technological Forecasting and Social Change by Professor Lujie Chen of International Business School Suzhou (IBSS) at Xi’an Jiaotong-Liverpool University (XJTLU) and her collaborators shows that GenAI does not automatically help organisations define problems, interpret user needs or translate outputs into feasible innovation pathways. The real challenge is not whether to adopt GenAI, but how to combine it with human‑centred design thinking.

The study examines how three GenAI capabilities — relational capability, which supports coordination and communication; analytical capability, which supports data processing and insight generation;, and creative capability, which supports the generation of new  content — work together with five  dimensions of design thinking: user focus, problem framing, visualisation, experimentation and iteration, and embracing diversity.)

Using survey data from 303 Chinese firms and applying fuzzy-set qualitative comparative analysis (fsQCA), the research finds that high levels of both types of innovation are achieved only when firms combine GenAI capabilities with design thinking, rather than relying on either one alone. Relying on either technology or design thinking alone is unlikely to deliver the desired results.

The study further shows that the two types of innovation  require different capability configurations. High exploratory innovation is associated with analytical capability, creative capability, problem framing, and embracing diversity. This suggests that firms need not only  GenAI's ability to process information and generate new ideas, but also the ability to reinterpret problems and bring in a wider range  of perspectives.

High exploitative innovation, by contrast, is associated with relational capability, user focus, and visualisation. This type of innovation relies more on  coordination, communication and the ability to refine existing products in close alignment with user needs. The findings suggest that the capabilities needed for one type of innovation cannot  simply be transferred to another.

For managers, the implications are clear. Firstly, firms should treat GenAI and design thinking as complementary resources rather than separate initiatives. Secondly, they should align  capability development with the type of innovation  they want to achieve:  exploratory innovation emphasising analytical and creative capabilities along with problem framing, while exploitative innovation requires relational capability and user focus. Thirdly, firms should translate these combinations into practical routines, such as cross‑functional design workshops, rapid prototyping, and real‑time feedback loops.

Professor  Chen says: "GenAI is not a universal innovation engine. Companies often invest heavily in improving AI capabilities but neglect design thinking, which acts as the steering wheel. No matter how powerful the technology's engine is, without direction it won't get you where you need to go."

About the author

Lujie Chen is a full Professor of Management at Xi’an Jiaotong-Liverpool University. Prof Chen was included in the Elsevier-Stanford University World's Top 2% Scientists 2024 and 2025. She is a Fellow of the Higher Education Academy in the UK and an expert in the fields of supply chain management and business analytics.

Professor Chen has published more than 60 papers  in top-ranking journals such as the Journal of Operations Management (UTD 24), Harvard Business Review (FT50), International Journal of Operations and Production Management (ABS 4), British Journal of Management (ABS 4), and European Journal of Operational Research (ABS 4), among others. She has served as a guest editor for special issues of several respected journals such as International Journal of Operations and Production Management, Industrial Marketing Management, International Journal of Production Economics, and Journal of Business Research. She is currently serving as an Associate Editor for the International Journal of Operations and Production Management (ABS 4) and department editor for IEEE TEM (ABS 3) and editorial board for Humanities and Social Sciences Communications (Nature Portfolio, CAS Humanities Q1 & JCR Q1).

Technological Forecasting & Social Change (TFSC) is committed to publishing research with a clear technological focus that makes significant contributions to both theory and practice. Technological innovation can optimize business activities, expand markets, and address social and environmental challenges. The journal covers four key areas: technology forecasting, managerial decision-making, impact assessment, and governance frameworks.