Hi,
In that case you don’t need to try to get a Pawn owner that doesn’t exists.
I had to ask you this since the begining, I’ve put you in the wrong direction sorry 
So,
you have a character and a trigger box (manually placed in the level) wrapping it, you are controlling a camera and when you overlap the box, you want the character in it look at the camera?
In your case, with only one Animation Blueprint, how your Event Graph is aware of the trigger event? It was aware because the logic was put in the Level Blueprint. But now? How the Anim Event graph knows when to start computations? (Maybe you can add a custom event in your event graph and call it from you characterMale node in the Level Blueprint)
Usually it’s better to have a Class Blueprint to wrap your mesh and a AnimBlueprint linked for skeletal related animation when apply. In your Class Blueprint you can attach a lot of different components to your mesh, like a trigger box for example. In such setup you can use PawnOwner node to get the Blueprint instance in the linked AnimBlueprint.
Anyway, lets’ try to have the animation working for instance, as the logic there should still be valid.
First, if you want to see something happening in your current event graph, you should pin the output of the SET ActorValue in the input of SET RotValue.
These pin are executions flows controllers, no pin, no party 
So,
In your AnimGraph you can remove the ActorValue node. The rest should be ok. The base pose is the head and you’ve choosen to weight half the rotation for the neck.
In your Event Graph you can remove all the Cast flow, you don’t need that.
You just need to find the location of the camera and the character.
The CameraManager Node can be used for anything related to the Camera.
For the character, it’s better to retrieve the location of the head bone to have a better target vector because ActorLocation is the Root location of the Object and in the case of a character, I don’t think the root location is the best one regarding aiming 
Try that to see if it works.
If it’s the case, you’ll probably notice that the head is facing the camera without any feeling of animation at all. This is the moment to introduce the time factor and interpolation as well.
RInterp Node, that’s it.
You need to pass the time elapsed since last frame, which is always given in the output pin of the AnimUpdateNode.
You have to choose the speed of animation, hence the interpolationSpeed pin that should be the degrees/second you want your head rotate.
There is also a RInterpToConstant Node with the exact same parameters.
Play with it to see the differences in the animation between both nodes.