No, the predictors making that prediction weren't doing propaganda, and most mainstream outlets didn't make that (or any other prediction), though many reported the fact of the existence of one or more of the predictions that were made.

Those making the 95% prediction were using statistical models established for that purpose before the identity of the candidates or what those models would show were known. They quite arguably are bad models as they assumed state level deviations from polling results are independent where history suggests that, in fact, they are strongly correlated (a fact pointed out by Nate Silver prior to the election, in explaining why 538 had a much lower projection of Clinton's probability of victory.) But choosing a bad model isn't propaganda.