Abstract
Generative artificial intelligence tools such as ChatGPT have created new opportunities for language learning. In this study, the author explores Saudi university English as a foreign language students' perceptions of ChatGPT as a digital learning companion, drawing on a technology-acceptance perspective to examine its usefulness, reliability, and challenges. Data were collected through a self-reported cross-sectional online questionnaire of 17 Likert-scale items adapted from validated measures across three domains. A total of 618 students from varied disciplines participated. Associations were assessed using multivariable linear regression. Findings revealed 63% to 75% agreement on usefulness items, showing strong internal consistency (α = .93). Students were more cautious regarding reliability and challenges. Perceived challenges showed no meaningful association with usefulness (r = .003, p = .940), indicating students valued ChatGPT regardless of difficulties. Overall, students held largely positive perceptions of ChatGPT's usefulness as a supplementary tool while remaining aware of its reliability limitations.