Maximum likelihood estimators converge in probability to the true values (are consistent) as long as certain regularity conditions hold (see https://en.wikipedia.org/wiki/Maximum_likelihood_estimation).
