Implicit in the Turing test is the entity doing the evaluation. It's quite possible that a human evaluator could be tricked, but a tool-assisted human, or an AI itself could not be. Or even just some humans could be better at not being tricked than others.