When and why do users trust AI in the kitchen? A hybrid modelling approach to the adoption of AI-assisted cooking

Journal article

This study explores how users perceive AI in various culinary scenarios. Using a hybrid model, we examined how perceived AI accuracy, general attitudes toward AI, and AI anxiety shape trust in AI cooking applications, and how this trust influences adoption intentions. Data from 380 UK participants revealed two groups: a small segment of Engagers, open to AI-assisted cooking, and a larger group of Avoiders. Trust was positively influenced by perceived accuracy and general attitudes, and negatively by AI anxiety. Trust, in turn, was key to distinguishing Engagers from Avoiders. However, adoption was highly context-dependent: AI was more accepted in practical tasks, such as using leftovers, but less so in socially meaningful settings such as dinner parties. Demographically, younger, educated individuals were more receptive. These findings highlight both opportunities and barriers for AI in the kitchen, offering insights into consumer trust and the nuanced dynamics of AI adoption in food contexts.

Authors: Califano, G; Zhang, T; Spence, C

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Taylor & Francis
• Journal: International Journal of Human-Computer Interaction
• Volume: 42
• Issue: 1
• Pages: 131-143
• Publication date: 2025-05-26
• Acceptance date: 2025-05-07
• DOI: 10.1080/10447318.2025.2505154
• EISSN: 1532-7590
• ISSN: 1044-7318
• Language: English
• Keywords: human–computer interaction; PLS-SEM; finite mixture modelling; technology acceptance; consumer behavior