The view of Kant was that because of how unpredictable the consequences of our actions were, we should not be held responsible for these consequences. He constructed a system of morality instead based on universal categorical imperatives that represent general rules that in theory, should be followed because they are consistent and non-contradictory. History has, aside perhaps from John Rawls efforts in A Theory Of Justice, moved on since then, and few people can be considered proper Kantians.
However, the deontic thinking of Kant is still respected as having been an important milestone in the development of morality. And in some sense, there is still value in the ideal of the universal categorical imperative. The reason is thus. Even within a more recent, consequentialist framework, there is still the reality that some consequences are uncertain.
So, even if one is a Utilitarian, it is still beneficial to consider having what I would call, Kantian Priors. The idea is simply that given uncertainty, we may not be sure what the right thing to do is, so we should err on the side of universal principles such as categorical imperatives to allow us to have a default position to take before any case specific information or evidence is processed. The ultimate posterior actions that we take are thus a result of a kind of Bayesian probabilistic formulation, where we take these priors, such as Tell the Truth, as our virtuous bias or prior probability, and then modify or update this with data that might show that Telling the Truth in this particular case, may not be the best thing to do. In this way we err on the side of universal principles, and mitigate the risk of doing evil unnecessarily due to inadequate information about the particular circumstances.
These priors do not necessarily have to be Kantian. They could also conceivably be based around Virtue Ethics, or even Christian or other religious moral principles if the practitioner finds these more compelling. The point is to have stable default positions that provide a degree of consistency to one’s moral reasoning and behaviour across different situations.
A part of the benefit of this is that such consistency is more understandable and reliable than purely consequentialist methods, which can sometimes be seen as overly calculating and justifying anything in practice. Thus, the addition of Kantian or other deontic priors, provides a kind of robustness to the moral framework, without making it so overly inflexible as to be unable to deal with contingencies.