The MASSIVE dataset and the Massively Multilingual NLU (MMNLU-22) competition and workshop will help researchers scale natural-language-understanding technology to every language on Earth.