| Although digital learning (DL) and self-regulated learning (SRL) have been widely examined independently, their interrelationship, moderating influences, and combined effects on L2 learning remain underexplored. This meta-synthesis addressed this lacuna by analyzing 56 high-quality empirical studies (N = 12,753) identified through a systematic search of JCR- and Scopus-indexed journals. Using thematic analysis in NVivo 12 and Sandelowski and Barroso’s synthesis framework, the findings were synthesized into five overarching themes. First, DL environments frequently enhanced SRL by supporting autonomy, monitoring, attention, and learner responsibility. Second, SRL–DL interactions were shaped by learner-related factors (motivation, self-efficacy, digital literacy, attitudes) and educational conditions (scaffolding, interaction, engagement, and facilitating conditions). Third, technology-supported SRL commonly involved goal-setting, self-monitoring, self-evaluation, and time/resource management, whereas social strategies were least used. Fourth, technology-based self-regulated learning (TBSRL) consistently improved language learning outcomes, particularly in vocabulary and writing, with several interventions showing sustained achievement gains. Finally, DL offered benefits such as increased motivation, personalization, and reduced anxiety, while concerns included overreliance on AI, academic integrity, and the need for human mediation. This synthesis provides important implications for theory and practice by proposing an integrated model of TBSRL in which digital affordances influence L2 outcomes through SRL processes moderated by learner and educational factors. The findings further highlight that effective AI- and technology-mediated language learning depends on access to digital tools, sustained pedagogical scaffolding, and support for collaborative regulation. |