While the discovery rules generally allow for liberal discovery, when it comes to eDiscovery, and a robot or A.I. would clearly fall under that hub, courts are less inclined to allow unrestricted access to an adversary's system, especially if there are no reasons to believe they've been hiding anything. It's almost always going to be a question of proportionality. One of the unanswered questions that a court will soon be faced with is whether a robot can be deposed, or actually take the witness stand. While clearly there is another way to extract data from a robot and present that data, when the merger of A.I. and robotics actually produces functional, Westworld-quality robots (or even just close to it), how will the courts deal with robot discovery? How will the court, and parties, ensure robots are telling the truth when queried? Rule 601 presumes that witnesses are persons.