In the same circumstances a light beam stops producing an interference pattern in the Young experiment, quantum wave functions do as well. This is pretty easy to derive, just introduce a random phase shift term, and average across it, and the interference pattern disappears and a bell curve emerges instead.
A state is entangled when it isn’t a product state.
Two spin (1/2) particles in a singlet state have the kind of “they have opposite states” thing going on that you describe, and is a specific way that two particles can be entangled.