If person A can't accept or understand that a human life overrides lesser considerations, then no, I don't put my trust in them.
"I'm an empath." "I don't like drama." "I never cheat people." "I value honesty."
Whut?
Surely it would also depend on the situation, and the relevance and reasoning behind B's view.
Are we in preschool with children? Then probably A is right.
But if B is a teacher and explains that the kids love a game in which they all rampantly cheat, and the teacher has given up because they are having an absolute blast breaking the rules and trying to trick each other? I hope you would change your mind too.
Are we talking about an undercover agent in a dangerous country, attempting to get a critical component from a drunk bioweapons scientist, at a card table in a casino?
These are humorous examples, but real world versions are not hard to come by.
Principles that have few or no exceptions tend to be very narrow in scope. Like don't preemptively launch world ending nukes during a stable peacetime.
The sensible approach is have the best principles you can, be willing to improve them, and apply them with care and situational flexibility.
Principles are maps, not the actual moral territory.
Principles are wisdom, not an algorithm.
The easiest retort is Anne Frank. You're hiding her in your attic and a Nazi asks if you're hiding enemies of the state. There you go, a time when you'll definitely lie and cheat!
Someone might answer, "well, fine. I don't cheat or lie unless I'm in extraordinary circumstances." That's fine, they've let go of dogmatism then, now the interesting conversation starts of where the line is, what constitutes extraordinary circumstances. That's a very interesting conversation I believe.
I don't think this is the win you think it is. Kantians and se deontologists will absolutely say that no, you cannot lie and cheat even in that scenario. You have a moral duty to not lie but also a moral duty to resist tyranny. You cannot sacrifice one to achieve the other, you must choose only options that fulfill both duties.
The universe doesn't respect that viewpoint. There is no mechanism in reality or life that prevents hard tradeoffs from having to be made.
What the universe does or doesn't respect has no bearing on what is or is not right / good.
That would be good and right, indeed.
Or, making actual choices in reality? Where there are limits to what we can do, but making hard choices well has positive impact.
I am speaking to the latter.
If a Kantian can be put into a situation where their morals would require them to say the word that gives up Anne Frank, we can safely say it's a bad moral system.
Not unless you can present some proof of this. Your implicit assumption that we should care about outcomes over principles has its own set of moral failures, like the repugnant conclusion.
And I maintain my simple point: if your ethical system doesn't allow the flexibility to not give up Anne Frank, it's a bad ethical system. Unless you believe giving up Anne Frank isn't wrong? Then you're a bad person and shouldn't be considered in conversations about ethics!
Design it in a way to have good outcomes if you're worried about repugnant conclusions. Personally I believe putting it on paper is a fool's errand - vibes based ethics seems to work as good as one can get from an ethical system.
In your hypothetical situation, I owe no such obligation to the Nazis who as you'll remember were an occupying force. I entered into no social compact with them.