I still dont trust the Anthopic and OpenAI are not training on my code. I even just thinking keeping track of what code you have received in prompts and to train/not train on it seems like an impossibly difficult task.