AimsThis meta‐analysis investigated the efficacy and safety of fully closed‐loop automated insulin delivery (AID) in patients with type 2 diabetes.Materials and methodsWe systemically searched PubMed, Scopus, Web of Science, and Cochrane Central. We included randomized controlled trials (RCTs) comparing fully AID versus conventional insulin therapy. The outcomes were pooled as Hedges' g and risk ratio (RR) in the random effects model with 95% confidence interval (CI). Our primary outcome was the proportion of time in the target glucose range. Key secondary outcomes included the proportion of time in hyperglycaemia or hypoglycaemia.ResultsWe included seven RCTs, compromising 390 patients. Our analysis showed that the fully closed‐loop AID increased the proportion of time spent within the target glucose range compared to the control group by additional 18 minutes per 24 hours (Hedges' g= 1.22%, 95% CI [0.84%, 1.6%], P <0.01), additional five minutes overnight (Hedges' g= 0.99%, 95% CI [0.65%, 1.33%], P <0.01), and additional 13 minutes during the daytime period (Hedges' g= 1.36%, 95% CI [1.1%, 1.61%], P <0.01). Compared to the control group, the overall time in hyperglycaemia was shortened by 15 minutes per 24 hours (Hedges' g= ‐1.04%, 95% CI [‐1.43%, ‐0.64%], P <0.01). There was no significant difference between the two groups in terms of overall, overnight, and daytime periods spent in hypoglycaemia.ConclusionsOur meta‐analysis suggested that fully closed‐loop AID may improve glycaemic control in patients with type 2 diabetes. Further research is required to establish the feasibility of implementing these systems in clinical practice.