am i right in assuming your code is closed-source?

i'd expect anything on github for example to be already in their training set or is training on actual usage more useful to them?