I wonder if by that token you could find a security exploit in an API that causes undesired behaviour (e.g. elevated privileges) and claim you're simply accessing an undocumented API?
Or another way of looking at it - does accessing an undocumented API constitute hacking/unauthorized access? (which is probably an even more serious violation than copyright infringement in most countries)
(disclaimer: I don't even know what this particular API is doing, or what's the alleged infringement, I'm just wondering about the principles in general)