A recent discovery has sparked concerns among experts about the training data used by OpenAI to develop its popular character, Sora. It appears that the AI was trained on a vast amount of game content, which has raised questions about potential copyright infringement and the implications of using such data without proper authorization.