Technically, that's true - our model of universe requires no fairies to exist so we have to assume they don't (otherwise where do we stop - everything imaginable has just as much likelihood of also existing).
If an actual fairy is discovered tomorrow and passes all reasonable scientific standards of evidence, the model was wrong and what we reasonably held to be "true" turns out not to be. Only a bad scientist would go on maintaining their previous "truth" when the evidence contradicts it (and the gold standard would be that a new model of the universe that includes fairies is able to make better predictions than the previous one).