Since its very beginnings, AI has more or less developed in parallel with two lines of research method paradigms. The first paradigm could be called statistical (pattern recognition, machine learning, also deep learning, which in the last few years has entered quite a vibrant phase of development). The second paradigm can be called logical. It (mostly) deals with automatic deduction systems and tool development in the environment of corresponding formal methods, which in particular encompass formal logic calculi. This paradigm partially uses these systems for the requirements of modelling and solving problems from the AI domain. There are many logic calculi that have found their application in modelling and solving a wide range of AI problems. These range from classical propositional calculi, their fragments (such as calculi of functional and multivalued dependencies, without which the relational data model wouldn't be possible), intuitionistic propositional logic and its many fragments and variants, superintuitionistic logics, multiple valued logics (Lukasiewicz logics), discrete as well as continuous, systems of modal propositional logics, first order predicate calculus (logic) and its variants, second order predicate Logic, F-Logic etc. In addition to these, we cannot avoid mentioning a whole spectrum of contextual domain logics such as fuzzy logics. Many of the logic calculi mentioned above have themselves become foundations for building logic programming languages such as Prolog (and its relatives), hybrid programming languages and tools. Next to the logic component, they encompass classical linear programming (constraint logic programming languages) and also specialized tools such as SAT-solvers, languages that implement 2 nd order predicate logic (HiLog) or tools such as Coq. Coq is based on a fragment of lambda calculus that for the last thirty years has been developed by INRIA, the world renowned computer science institute based in France. Through this presentation of selected logic systems, related programming languages, development environments and specific purpose tools, the author wants to highlight the enormous potential of their application in modelling and solving AI problems.