Generative AI framework for sensory and consumer research

Generative artificial intelligence (GenAI) technologies, including ChatGPT, offer innovative capabilities in sensory and consumer science. Recent empirical studies in sensory and consumer science highlight the potential utility of GenAI in, for example, AI-generated food images and recipes. To the best of our knowledge, this is the first paper to propose a comprehensive framework for integrating GenAI into research and development in sensory and consumer science. The framework highlights how GenAI can be applied across the concept, design, and testing phases through an iterative process. The concept phase utilises GenAI to generate research concepts (e.g., proposing ideas such as research questions and hypotheses). The design phase employs GenAI to formulate research designs. During this stage, GenAI assists with creating and validating survey/experimental stimuli and measurement scales. The testing phase applies GenAI to evaluate research ideas and designs by employing “silicon samples,” interactive surveys that enhance engagement and response quality. In the testing phase, GenAI can also analyse unstructured text data, offering more accurate and scalable text analysis than traditional methods, even across diverse languages and cultures. This study also acknowledges potential pitfalls, such as biases in AI outputs, data privacy and security concerns, oversimplification, lack of transparency, and GenAI user misperception. This article encourages greater integration of GenAI by highlighting its potential for the sensory and consumer science community, while addressing its limitations and ensuring adherence to high ethical standards.