What Does the Legal Gamut Look Like?

We're seeing a whole host of new challenges rise up as we're starting to build these AI systems. We see already debates in the US about law enforcement using facial recognition. Who owns that art? That's going to be the case for both images and for text. And who owns some new invention that an AI system comes up with? What if you use the AI to do that? How much do you have to modify it to have ownership over it? It's also an issue when we come to think about the data that these AI systems are trained on. There's they're trained right now on copyrighted data. But that's an open question.