With the increase use of artificial intelligence methodologies even in safety-critical systems for implementing automated and autonomous functionality, there is a need for coming up with verification and certification methodologies allowing to justify the degree of verification at least to a certain extent. Combinatorial testing makes use of the combinatorial strength for this purpose. In recent work, we reported on an extension of combinatorial testing, i.e., ontology-based testing, where we utilize ontologies for describing environmental models that can be mapped to combinatorial testing input models. In my talk, I introduce the basic ideas, discuss the foundations, and present results obtained when using combinatorial testing from various domains, including autonomous and automated driving functions and security testing.