Mr. Seán O'Sullivan:

All of the data sets I used are in the public domain. I am exploring training a larger version using the Project Gutenberg corpus, which comprises 70,000 books that are all in the public domain. I have not explored or considered using potentially copyright works and I do not plan to.