This study investigates the influence of the automated insulin delivery system (AID) on glycaemic control in children and adolescents with type 1 diabetes (T1D) who do not reach optimal glycaemic control with traditional treatment options. All the patients aged 7 to 16 years with T1D who initiated the AID system between 24 October 2020 and 5 January 2022 in the Helsinki University Hospital and had haemoglobin A1C (HbA1c) levels above 53 mmol/mol/7.0% (N = 79) were included. Time in tight range (TITR), time in range (TIR), HbA1c, mean sensor glucose (SG) value, time below range (TBR) and SG coefficient of variance (CV) were measured at 0, 3, 12 and 24 months. The changes in the outcome measures between the time points were included in the analyses, and statistically significant level was p-value <0.01. After the initiation of AID, glycaemic control improved, and the effect lasted throughout the study period. Between 0 and 3 months, TITR and TIR increased (mean 11.7% [10.6], mean 18.1% [standard deviation [SD] 13.7], p < 0.001), whereas HbA1c and mean SG values decreased significantly (-8.3 mmol/mol [8.7]/-2.9% [2.9], p < 0.001, -1.8 mmol/L [1.7], p < 0.001). These effects were sustainable and were still visible at 12 and 24 months. Glycaemic control in patients not reaching treatment goals improved significantly after the initiation of the AID system, and the favourable effect lasted throughout the follow-up. AID treatment could be an option for also those paediatric patients with T1D who do not have good skills in diabetes management.