Google, OpenAI and the Coming Copyright Storm

So here is a question: If you develop some software and train it on random material from the internet to make it smarter, do the creators of that material have any right to the software?

What happens if you are Google and you train your software with public conversations generated from a chatbot made by OpenAI, one of your main competitors, which has been hiring your top talent? Is that smart? Sketchy? Somewhere in between?