I think it would be unreasonable to claim that it is fake anyway, since relevant research has been public for some time. If there's an amount of trickery involved, it's in the parameter files he loaded (they might be nontrivial/nonautomatic to produce).