["British Journal of Educational Technology, EarlyView. ", "\n\nSelf‐directed learning (SDL) is pivotal for academic success and lifelong learning, in which motivation has been a major challenge, especially in the context of self‐directed out‐of‐class reading. According to self‐determination theory (SDT), the fulfilment of three basic psychological needs (BPNs)—autonomy, competence and relatedness—is key to autonomous motivation. However, it is hard to provide personalised support to BPNs in self‐directed out‐of‐class reading. Generative artificial intelligence (GenAI), with its capacity for generative human‐like outputs, makes the provision of personalised interactive support possible. This study thus examined how GenAI‐powered chatbots can be utilised to serve as reading companions to fulfil BPNs in self‐directed out‐of‐class reading. An online reading platform was developed, where students engaged in self‐directed out‐of‐class reading. Sixty‐nine students from two classes at a Chinese university were randomly assigned to two reading conditions: the experimental group (EG; N = 36) where students had access to a GenAI‐powered chatbot as a reading companion on the platform; and the control group (CG; N = 33) where students did not have access to the GenAI‐powered chatbot during reading. Multiple data sources were collected and analysed, including pre‐ and post‐reading tests, pre‐ and post‐questionnaires on reading motivation and self‐directed out‐of‐class reading, log data on reading amounts and semi‐structured interviews. The findings revealed that EG showed significant improvement in autonomous reading motivation and self‐directed out‐of‐class reading. Reading performance did not differ between the two groups. This study offers valuable evidence on the efficacy of integrating GenAI chatbots into SDL through the lens of SDT. Pedagogically, educators can leverage GenAI technology to cater to students' BPNs, bolstering their SDL capabilities and academic achievement.\n\n\n\n\nPractitioner notes\nWhat is already known about the topic\n\nSelf‐directed learning (SDL) has been recognised as a crucial competence for academic success and lifelong learning.\nMotivation has been a major challenge in self‐directed out‐of‐class reading, with the fulfilment of basic psychological needs (BPNs) identified as vital for autonomous motivation according to self‐determination theory (SDT).\nGenerative artificial intelligence (GenAI) has the potential to fulfil students' BPNs by providing personalised interactive support to enhance their motivation and SDL.\n\nWhat this paper adds\n\nThis study adopted the theoretical lens of SDT to develop a GenAI‐empowered online reading platform with a companion chatbot named Reade, aiming to fulfil students' BPNs in self‐directed out‐of‐class reading.\nThis study examined the efficacy of the designed chatbot‐assisted BPNs (CaBPNs) in supporting self‐directed out‐of‐class reading.\nThis study offers empirical evidence that CaBPNs support can enhance students' reading motivation and self‐directed out‐of‐class reading.\n\nImplications for practice and/or policy\n\nEducators should capitalise on the potential of GenAI chatbots in fulfilling learners' BPNs through the provision of personalised interactive support to boost autonomous motivation and SDL.\nEducators should be aware of the role of AI literacy in leveraging the full potential of chatbot support for SDL and take proactive steps to cultivate learners' AI literacy.\n\n\n\n\n"]