Yes, there is a possible problem here. I have had that problem myself. Suggest just to be careful, no more analysis on the chess machine for now as it might lead to a future position.
I generally won't analyze any ongoing game position, that has had 10-15 or more moves played in it give or take (if it's very early in the opening and there are many options this is more of a possible grey area), that could directly arise from it. So for example, I wouldn't analyze a position that came about by me playing out (by myself) 5 moves, move-by-move, from a position I obtained from an ongoing game I am playing. That seems like a good way to err on the side of caution.
I'm not sure on the exact rules, as that would be hard to determine, but I do think it's courteous to try hard to not make the mistake of looking at computer analysis of a position that will most likely occur soon in one of your games. Or at least don't do so deliberately, but even though we can't always not make accidental mistakes, we can make them much less likely if we take a cautious approach.
If of course it's really that important to you to analyze a certain line, just resign the on-going game to ensure that you are not making use of an unfair advantage. It's better that the analyzing of the position comes at your expense rather than your opponent's.
I agree with you and have been doing exactly that for the same reason, I just wanted another opinion.