A quick Google search on the subject points to a reference that does not suggest molecular sieves for drying ethylene glycol. I don't know if you have a reference that suggests it, or if this is a new occurrence with a previously used procedure, but you should have some source of suitability for this task.http://delloyd.50megs.com/moreinfo/drying.html
You say there was more moisture after the procedure, perhaps your sieves were old and saturated with moisture? But you felt the reaction with water, so that must not absolutely be the case. Still, it could have some water, and still be able to absorb more.
The cloudiness, if it is pronounced, could be the sieves dissolving in the ethylene glycol. Or it could just be the refractive index of moisture dissolving. I didn't see what you saw, so I can't differentiate the two. I don't know if ethylene glycol dissolves molecular sieves, or if ethylene glycol can enter a 3A pore. If it can, then it will likely drive the moisture out, and ruin the bulk of the solution.
Sometimes old molecular sieves just crumbles, and contaminates your solvent, so maybe that's what happened.
I've performed Karl-Fisher titration of very complex mixtures, so I'd suspect the cloudiness wouldn't cause a false high reading. But I'm not sure.
Sorry, but that's an awful lot of "I dunno". Can you try again and see if you get the exact same problem? Have you done this before so you're sure this works? Can you try again with fresh reagent to see if this is a fluke?