Not in a sense, like humans do, like lying to hurt someone, or lying not to hurt someone (White lies), but lying because it was trained on data humans produced, who sometimes, lie for whatever reasons.
Imagine the trained is like: A->B-C->D (humans typically lie on answer B) the NN picks it up, and lies too.
Imagine the trained is like: A->B-C->D (humans typically lie on answer B) the NN picks it up, and lies too.
Just a thought I just had :)
Maybe not this generation, but in two?