In Frankenstein's defense I would argue that the monster has a conscience that's equivalent to an adult human conscience (at least when some of the crimes were committed, if I remember the book correctly), and so the monster is capable of legal guilt. In that scenario, Dr. Frankenstein is akin to the monster's parent. We can decry his actions but, ultimately, it's the monster who's responsible.

In prosecuting Frankenstein I'd argue the opposite: That the monster is merely a machine (albeit a biological machine) and therefore has no conscience. Granted, that argument would have worked better in the late 18th century than today!

What's interesting to me about this question is that it probably won't be much longer before it goes from being hypothetical to being an actual legal case. Presumably, at some point in the not too distant future, a human-created machine with at least the appearance of sentience will harm someone and we'll have to decide, legally, who's responsible. I don't think we know how to determine whether such a machine has a conscience. After all, the consciousness of a human defendant is only a legal presumption. We have no way of knowing it exists. Will we decide to extend that presumption to anything that behaves like it has consciousness? Anything that claims to have consciousness?