Introducing students to a new law is pedagogically challenging. One can just declare it as fact, but it is easier on (at least some) students if we can derive it when possible, or, if not, at least motivate it somehow. The Biot-Savart law is particularly challenging in this respect, because it is only derivable from more advanced formulations (such as Maxwell’s equations), which are in their turn merely stipulated (this is the approach of Feynman et al., for example1). The other possibility is simply to declare its empirical truth, without details (see, for example, Griffiths2).
Ultimately, any new law or relation is founded on experimental results, but we should still strive to use the simplest possible empirical input, that which can be assimilated and accepted most easily. For example, when teaching electricity in introductory courses, Coulomb’s law is taken as the fundamental empirical relation, rather than...