Ask HN: Is there utility in NLP today?

Given today's state-of-the-art in NLP, is it possible to semantically analyze the GDPR corpus in order to create a human-readable decision-tree that would give businesses a way of validating their compliance with said law?

Kind of like:

----------------------

Start. Do you show a "do you constent to blah blah blah"?

No. Are you tracking stuff?

Yes. Does it contain PII?

No. You're good.

----------------------

Is it possible? How would you go about producing it?