Before I can begin to answer the question at hand, it is useful to first clarify the key terms iterated. By Deontological, I am referring to ethical theories that place a great focus on motivations that drive us to act in a certain way, regardless of the direct consequences of the action. The most prominent deontological ethical theory is that of Kant's 'Categorical Imperative'. The theory states that you should only act if you could will the action to become a universal law. Essentially he is suggesting that if we could not conceive of everyone being able to perform the action at any given point, then we should not perform the action. If we take an action such as murder, it is clear to see that we would not want everyone to perform the action, as it would put yourself and those you loved in danger of being murdered without repercussions.
While this theory is seemingly in line with our intuitions regarding morality, (i.e. murder as an act is wrong) the theory itself struggles in some circumstances. One of these circumstances arises when you have a clash of two actions that you wouldn't want to be universalised. To help illustrate this problem, I will use the famous example of an 'Axe Murderer' knocking at your door. In the example, an Axe Murderer knocks at your door asking if your family are in (with the intention to kill them). As your family are in, you would be forced to lie to the murderer in order to protect them. This poses a grave dilemma for the Kantian ethicist, as telling a lie is an action that you would not want to be universalised - if everyone lied then it would diminish the value of the truth. At the same time, it is clear that you don't want the Murderer to do what he does best and murder your family. It seems as if the Categorical Imperative falls short at this hurdle and cannot provide an answer. On the contrary, it could be argued that it is not lying in general which you are willing to become a universal law, instead it is lying to protect your family - something that most people would be inclined to universalise. Unfortunately for the Kantian this also serves as problematic for their theory. This is down to the fact that if you allow certain amounts of refinement of the action (changing lying to lying for your family), it opens you up to a potential fallacy known as an infinite regress. You could always add more specific criteria to the action in order to avoid problems with it being universalised which lends me the belief that you could manipulate any action into being permitted by the Categorical Imperative. Overall therefore, the Categorical Imperative gives us a platform that helps us decide how we should morally act. This platform can be opened up to manipulation however, which subsequently grants people the opportunity to act as they please and claim that it is 'Morally Permissible'.