Answers are grounded in this paper and cite their sources.
Meta-analysis evaluates AI chatbots and self-help apps, showing moderate symptom reduction in social anxiety.
Due to stigma, cost, and a lack of trained specialists, social anxiety disorder (SAD), one of the most common and debilitating mental health conditions, is still challenging to treat using evidence-based methods (Kessler et al., 2005). In recent years, scalable tools for managing mental health conditions like SAD have emerged, including self-guided mobile applications and chatbots powered by artificial intelligence (AI) (Fulmer et al., 2018; Inkster et al., 2018). This systematic review and meta-analysis assessed the efficacy of AI-based interventions, specifically chatbots and self-help apps, in lowering symptoms of social anxiety from 2015 to 2025. Following PRISMA guidelines, only peer-reviewed randomized controlled trials (RCTs) that measured social anxiety using validated measures such as the SPIN, LSAS, and SIAS were included. The results of quantitative synthesis using Hedges' g consistently supported digital interventions over control conditions, with a small to moderate pooled effect size (g ≈ 0.36, 95% CI [0.29, 0.43]) (Stolz et al., 2018; Ivanova et al., 2016). The resources included modular self-help apps built with CBT or ACT frameworks and AI chatbots like Woebot and Wysa. Even though the results point to significant short-term advantages, there are still a number of limitations, such as the absence of long-term monitoring, a small sample diversity, and low capacity for cultural adaptation. More beneficial for settings having limited resources as these tools show potential alternatives or substitutes for therapist-led care. Future studies should concentrate on ethical inclusion, culturally sensitive design, and comparative trials with traditional therapies.
No takes yet. Share an insight, caveat, or question.
Mushtaq et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: