BERT (Bidirectional Encoder Representations from Transformers) was developed by researchers at Google AI Language in 2018. The team included Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. BERT is a state-of-the-art pre-training language model that has significantly improved various natural language processing tasks. It is based on the Transformer architecture and uses a bidirectional approach to capture contextual information from both left and right contexts. This pre-training approach allows BERT to learn rich representations of words and sentences, which can then be fine-tuned for specific downstream tasks, such as question answering, text classification, and named entity recognition. BERT has since become one of the most influential models in the field of natural language processing.