How I demostrated X bar squared (sample mean square) is a biased estimator of mu (population mean)squared?

It is not a biased estimator of mu coz if you do the mean of the sample mean distribution,it will actually give the population mean,mu