Comment by indoordin0saur

Sure. That's what they say but does anyone actually trust what they say about where they are getting their data? They also had a legal requirement not to steal IP, they said they weren't and then it came out that both OpenAI and Anthropic were pirating mass amounts of media. When they said they weren't doing this they were knowingly lying. I'm quite certain that some (if not all big players) are retaining data from their customers despite explicit agreements not to do so. As a data engineer myself, I know how easy this is to do.