There absolutely is a difference, yes.
But if the answer to "how do I tell the difference?" is that I ask the entity who is making the request in the first place, we've now achieved full epistemic closure.
That is: if I don't know whether Omega tells the truth or not, and I don't know whether Omega has my best interests in mind or not, and Omega tells me to sacrifice a child, I probably wouldn't sacrifice the child. Would you?
More generally, there is a big difference between "what ought I do, if X is the case?" and "what decision will the decision procedure that I ought to implement make, given non-zero but uncompelling evidence that X is the case?" Thought experiments often ask the former, but the latter is more relevant to my actual life.
But if the answer to "how do I tell the difference?" is that I ask the entity who is making the request in the first place, we've now achieved full epistemic closure.
I assume she expected you to ask God, y'know, now, not immediately after something claiming to be Him appeared and ordered you to kill 'em all. (Presumably asking Him "wait, are you sure killing children is a good idea?" would be met with a "yes". Or a thunderbolt.)
Happy New Year! Here's the latest and greatest installment of rationality quotes. Remember: