You could argue that the URLs are not wrong, but that the error is committed by the reader in expecting the output to somehow be based in reality.

I could roll a dice to generate a URL and then say “but it’s a wrong URL!” I don’t think many people would blame the dice.