The COVID-19 pandemic has accelerated the adoption of eLearning tools globally, with over 210 countries affected by school closures and teacher absenteeism. As schools reopen, speech-to-text recognition technology can play a crucial role in preventing children from falling behind and helping them catch up to pre-pandemic levels. However, current voice recognition systems are not accurate enough for children's voices due to differences in pitch, patterns, and cadences compared to adults. To address this issue, companies like Speechmatics have developed more accurate technologies, such as self-supervised learning and Autonomous Speech Recognition, which can handle background noise and improve accuracy significantly. These advancements aim to reduce AI bias and provide a step-change in speech recognition, enabling the use of subtitles in classrooms to develop literacy skills and support children's language development.