>"For the AU-led research team, which focused on COVID-related misinformation, human subject matter experts categorized a dataset of tweets, flagging a message as misinformation if it promoted political bias, conspiracies, propaganda, anger, or racism."
What a terrifying broad set of criteria.