Evaluating the Propensity of Generative AI for Producing Harmful Disinformation During an Election Cycle